| Publish Date | Title | Authors | URL | Abstract |
|---|---|---|---|---|
| 2025-08-29 | A Single Subject Machine Learning Based Classification of Motor Imagery EEGs | Dario Sanalitro, Marco Finocchiaro, Pasquale Memmolo, Emanuela Cutuli, Maide Bucolo | Link | Motor Imagery-Based Brain-Computer Interfaces (MI-BCIs) are systems that detect and interpret brain activity patterns linked to the mental visualization of movement, and then translate these into instructions for controlling external robotic or domotic devices. Such devices have the potential to be useful in a broad variety of applications. While implementing a system that would help individuals restore some freedom levels, the interpretation of (Electroencephalography) EEG data remains a complex and unsolved problem. In the literature, the classification of left and right imagined movements has been extensively studied. This study introduces a novel pipeline that makes use of machine learning techniques for classifying MI EEG data. The entire framework is capable of accurately categorizing left and imagined motions, as well as rest phases, for a set of 52 subjects who performed a MI task. We trained a within subject model on each individual subject. The methodology has been offline evaluated and compared to four studies that are currently the state-of-the-art regarding the specified dataset. The results show that our proposed framework could be used with MI-BCI systems in light of its failsafe classification performances, i.e. 99.5% in accuracy |
| 2025-08-29 | Technical Development of Two-Photon Optogenetic Stimulation and Its Potential Application to Brain-Machine Interfaces | Riichiro Hira, Yoshikazu Isomura | Link | Over the past decade, techniques enabling bidirectional modulation of neuronal activity with single cell precision have rapidly advanced in the form of two-photon optogenetic stimulation. Unlike conventional electrophysiological approaches or one-photon optogenetics, which inevitably activate many neurons surrounding the target, two-photon optogenetics can drive hundreds of specifically targeted neurons simultaneously, with stimulation patterns that can be flexibly and rapidly reconfigured. In this review, we trace the development of two-photon optogenetic stimulation, focusing on its progression toward implementations in large field of view two-photon microscopes capable of targeted multi neuron control. We highlight three principal strategies: spiral scanning, temporal focusing, and three-dimensional computer-generated holography (3D CGH), along with their combinations, which together provide powerful tools for causal interrogation of neural circuits and behavior. Finally, we discuss the integration of these optical technologies into brain machine interfaces (BMIs), emphasizing both their transformative potential and the technical challenges that must be addressed to realize their broader impact. |
| 2025-08-29 | A universal animal communication tempo resonates with the receiver's brain | Guy Amichay, Vijay Balasubramanian, Daniel M. Abrams | Link | During fieldwork in Thailand we observed nearly identical frequencies of co-located flashing fireflies and chirping crickets. Motivated by this, we perform a meta-analysis and show an abundance of evolutionarily distinct species that communicate isochronously at ~0.5-4 Hz, suggesting that this might be a frequency "hotspot." We hypothesize that this timescale may have a universal basis in the biophysics of the receiver's neurons. We test this by demonstrating that small receiver circuits constructed from elements representing typical neurons will be most responsive in the observed frequency range. |
| 2025-08-29 | Quantile Function-Based Models for Neuroimaging Classification Using Wasserstein Regression | Jie Li, Gary Green, Jian Zhang | Link | We propose a novel quantile function-based approach for neuroimaging classification using Wasserstein-Fr\'echet regression, specifically applied to the detection of mild traumatic brain injury (mTBI) based on the MEG and MRI data. Conventional neuroimaging classification methods for mTBI detection typically extract summary statistics from brain signals across the different epochs, which may result in the loss of important distributional information, such as variance, skewness, kurtosis, etc. Our approach treats complete probability density functions of epoch space results as functional response variables within a Wasserstein-Fr\'echet regression framework, thereby preserving the full distributional characteristics of epoch results from |
| 2025-08-29 | Testing quantum markers of brain processes | Partha Ghose, Dimitris Pinotsis | Link | The emergence of the Dirac equation from a stochastic master equation suggests a profound link between relativistic quantum mechanics and underlying probabilistic descriptions of brain dynamics. In parallel, recent work has shown that the FitzHugh-Nagumo equations describing excitable neurons can be reformulated to yield a Schr\"{o}dinger-like equation with a novel Planck-like constant, indicating that neural noise may give rise to quantum-like dynamics. This paper brings these insights together to propose two novel neuroscience experiments aimed at detecting emergent coherence in axonal signal propagation and subthreshold oscillations. We suggest that stochastic interference effects in axon branching structures may reveal signatures of Dirac-type stochasticity. We also suggest that by measuring neuronal temperature and fluctuations we can detect quantum effects in brain oscillations. If successful, these experiments will provide experimental support for quantum markers of brain processes. |
| 2025-08-29 | Federated Fine-tuning of SAM-Med3D for MRI-based Dementia Classification | Kaouther Mouheb, Marawan Elbatel, Janne Papma, Geert Jan Biessels, Jurgen Claassen, Huub Middelkoop, Barbara van Munster, Wiesje van der Flier, Inez Ramakers, Stefan Klein, Esther E. Bron | Link | While foundation models (FMs) offer strong potential for AI-based dementia diagnosis, their integration into federated learning (FL) systems remains underexplored. In this benchmarking study, we systematically evaluate the impact of key design choices: classification head architecture, fine-tuning strategy, and aggregation method, on the performance and efficiency of federated FM tuning using brain MRI data. Using a large multi-cohort dataset, we find that the architecture of the classification head substantially influences performance, freezing the FM encoder achieves comparable results to full fine-tuning, and advanced aggregation methods outperform standard federated averaging. Our results offer practical insights for deploying FMs in decentralized clinical settings and highlight trade-offs that should guide future method development. |
| 2025-08-29 | Self-regulated emergence of heavy-tailed weight distributions in evolving complex network architectures | Jia Li, Cees van Leeuwen, Roman Bauer, Ilias Rentzeperis | Link | Brain networks continually adjust the weights of their connections, resulting in heavy-tailed distributions of their connection weights, a few strong connections among many weaker ones. At the same time, these connections undergo structural plasticity, forming a complex network topology. Although mechanisms producing either heavy-tailed distributions or complex topologies have been proposed, it has remained unclear whether a single mechanism can produce both. We consider homeostasis as the driving principle and propose a Hebbian inspired model that adaptively adjusts weights and rewires directed connections based on homeostatic dynamics. Without adaptive rewiring, weight adjustment alone still generates heavy-tailed weight distributions, as long as activity does not spread beyond locally neighboring units. However, when combined with adaptive rewiring, the homeostatic dynamics create a synergy that produces heavy-tailed weight distributions also for more extended activity flow. Furthermore, the model generates complex network structures that encompass convergent-divergent circuits similar to those that facilitate signal transmission throughout the nervous system. By combining adaptive weight adjustment and rewiring based on the same homeostatic dynamics, our model provides a parsimonious and robust mechanism that simultaneously produces heavy-tailed weight distributions and convergent-divergent units under a wide range of dynamical regimes. |
| 2025-08-29 | Discovering Semantic Subdimensions through Disentangled Conceptual Representations | Yunhao Zhang, Shaonan Wang, Nan Lin, Xinyi Dong, Chong Li, Chengqing Zong | Link | Understanding the core dimensions of conceptual semantics is fundamental to uncovering how meaning is organized in language and the brain. Existing approaches often rely on predefined semantic dimensions that offer only broad representations, overlooking finer conceptual distinctions. This paper proposes a novel framework to investigate the subdimensions underlying coarse-grained semantic dimensions. Specifically, we introduce a Disentangled Continuous Semantic Representation Model (DCSRM) that decomposes word embeddings from large language models into multiple sub-embeddings, each encoding specific semantic information. Using these sub-embeddings, we identify a set of interpretable semantic subdimensions. To assess their neural plausibility, we apply voxel-wise encoding models to map these subdimensions to brain activation. Our work offers more fine-grained interpretable semantic subdimensions of conceptual meaning. Further analyses reveal that semantic dimensions are structured according to distinct principles, with polarity emerging as a key factor driving their decomposition into subdimensions. The neural correlates of the identified subdimensions support their cognitive and neuroscientific plausibility. |
| 2025-08-29 | A Financial Brain Scan of the LLM | Hui Chen, Antoine Didisheim, Luciano Somoza, Hanqing Tian | Link | Emerging techniques in computer science make it possible to "brain scan" large language models (LLMs), identify the plain-English concepts that guide their reasoning, and steer them while holding other factors constant. We show that this approach can map LLM-generated economic forecasts to concepts such as sentiment, technical analysis, and timing, and compute their relative importance without reducing performance. We also show that models can be steered to be more or less risk-averse, optimistic, or pessimistic, which allows researchers to correct or simulate biases. The method is transparent, lightweight, and replicable for empirical research in the social sciences. |
| 2025-08-28 | Activity propagation with Hebbian learning | Will T. Engedal, Róbert Juhász, István A. Kovács | Link | We investigate the impact of Hebbian learning on the contact process, a paradigmatic model for infection spreading, which has been also proposed as a simple model to capture the dynamics of inter-regional brain activity propagation as well as population spreading. Each of these contexts calls for an extension of the contact process with local learning. We introduce Hebbian learning as a positive or negative reinforcement of the activation rate between a pair of sites after each successful activation event. Learning can happen either in both directions motivated by social distancing (mutual learning model), or in only one of the directions motivated by brain and population dynamics (source or target learning models). Hebbian learning leads to a rich class of emergent behavior, where local incentives can lead to the opposite global effects. In general, positive reinforcement (increasing activation rates) leads to a loss of the active phase, while negative reinforcement (reducing activation rates) can turn the inactive phase into a globally active phase. In two dimensions and above, the effect of negative reinforcement is twofold: it promotes the spreading of activity, but at the same time gives rise to the appearance of effectively immune regions, entailing the emergence of two distinct critical points. Positive reinforcement can lead to Griffiths effects with non-universal power-law scaling, through the formation of random loops of activity, a manifestation of the ``ant mill" phenomenon. |
| Publish Date | Title | Authors | URL | Abstract |
|---|---|---|---|---|
| 2025-08-29 | A Single Subject Machine Learning Based Classification of Motor Imagery EEGs | Dario Sanalitro, Marco Finocchiaro, Pasquale Memmolo, Emanuela Cutuli, Maide Bucolo | Link | Motor Imagery-Based Brain-Computer Interfaces (MI-BCIs) are systems that detect and interpret brain activity patterns linked to the mental visualization of movement, and then translate these into instructions for controlling external robotic or domotic devices. Such devices have the potential to be useful in a broad variety of applications. While implementing a system that would help individuals restore some freedom levels, the interpretation of (Electroencephalography) EEG data remains a complex and unsolved problem. In the literature, the classification of left and right imagined movements has been extensively studied. This study introduces a novel pipeline that makes use of machine learning techniques for classifying MI EEG data. The entire framework is capable of accurately categorizing left and imagined motions, as well as rest phases, for a set of 52 subjects who performed a MI task. We trained a within subject model on each individual subject. The methodology has been offline evaluated and compared to four studies that are currently the state-of-the-art regarding the specified dataset. The results show that our proposed framework could be used with MI-BCI systems in light of its failsafe classification performances, i.e. 99.5% in accuracy |
| 2025-08-28 | EEGDM: Learning EEG Representation with Latent Diffusion Model | Shaocong Wang, Tong Liu, Ming Li, Minjing Yu, Yong-Jin Liu | Link | While electroencephalography (EEG) signal analysis using deep learning has shown great promise, existing approaches still face significant challenges in learning generalizable representations that perform well across diverse tasks, particularly when training data is limited. Current EEG representation learning methods including EEGPT and LaBraM typically rely on simple masked reconstruction objective, which may not fully capture the rich semantic information and complex patterns inherent in EEG signals. In this paper, we propose EEGDM, a novel self-supervised EEG representation learning method based on the latent diffusion model, which leverages EEG signal generation as a self-supervised objective, turning the diffusion model into a strong representation learner capable of capturing EEG semantics. EEGDM incorporates an EEG encoder that distills EEG signals and their channel augmentations into a compact representation, acting as conditional information to guide the diffusion model for generating EEG signals. This design endows EEGDM with a compact latent space, which not only offers ample control over the generative process but also can be leveraged for downstream tasks. Experimental results show that EEGDM (1) can reconstruct high-quality EEG signals, (2) effectively learns robust representations, and (3) achieves competitive performance with modest pre-training data size across diverse downstream tasks, underscoring its generalizability and practical utility. |
| 2025-08-28 | Spatiotemporal EEG-Based Emotion Recognition Using SAM Ratings from Serious Games with Hybrid Deep Learning | Abdul Rehman, Ilona Heldal, Jerry Chun-Wei Lin | Link | Recent advancements in EEG-based emotion recognition have shown promising outcomes using both deep learning and classical machine learning approaches; however, most existing studies focus narrowly on binary valence prediction or subject-specific classification, which limits generalizability and deployment in real-world affective computing systems. To address this gap, this paper presents a unified, multigranularity EEG emotion classification framework built on the GAMEEMO dataset, which consists of 14-channel EEG recordings and continuous self-reported emotion ratings (boring, horrible, calm, and funny) from 28 subjects across four emotion-inducing gameplay scenarios. Our pipeline employs a structured preprocessing strategy that comprises temporal window segmentation, hybrid statistical and frequency-domain feature extraction, and z-score normalization to convert raw EEG signals into robust, discriminative input vectors. Emotion labels are derived and encoded across three complementary axes: (i) binary valence classification based on the averaged polarity of positive and negative emotion ratings, and (ii) Multi-class emotion classification, where the presence of the most affective state is predicted. (iii) Fine-grained multi-label representation via binning each emotion into 10 ordinal classes. We evaluate a broad spectrum of models, including Random Forest, XGBoost, and SVM, alongside deep neural architectures such as LSTM, LSTM-GRU, and CNN-LSTM. Among these, the LSTM-GRU model consistently outperforms the others, achieving an F1-score of 0.932 in the binary valence task and 94.5% and 90.6% in both multi-class and Multi-Label emotion classification. |
| 2025-08-28 | Towards Automated EEG-Based Detection Using Deep Convolutional Autoencoders | Annika Stiehl, Nicolas Weeger, Christian Uhl, Dominic Bechtold, Nicole Ille, Stefan Geißelsöder | Link | Epilepsy is one of the most common neurological disorders. This disease requires reliable and efficient seizure detection methods. Electroencephalography (EEG) is the gold standard for seizure monitoring, but its manual analysis is a time-consuming task that requires expert knowledge. In addition, there are no well-defined features that allow fully automated analysis. Existing deep learning-based approaches struggle to achieve high sensitivity while maintaining a low false alarm rate per hour (FAR/h) and lack consistency in the optimal EEG input representation, whether in the time or frequency domain. To address these issues, we propose a Deep Convolutional Autoencoder (DCAE) to extract low-dimensional latent representations that preserve essential EEG signal features. The ability of the model to preserve relevant information was evaluated by comparing reconstruction errors based on both time series and frequency-domain representations. Several autoencoders with different loss functions based on time and frequency were trained and evaluated to determine their effectiveness in reconstructing EEG features. Our results show that the DCAE model taking both time series and frequency losses into account achieved the best reconstruction performance. This indicates that Deep Neural Networks with a single representation might not preserve the relevant signal properties. This work provides insight into how deep learning models process EEG data and examines whether frequency information is captured when time series signals are used as input. |
| 2025-08-28 | Adaptive Segmentation of EEG for Machine Learning Applications | Johnson Zhou, Joseph West, Krista A. Ehinger, Zhenming Ren, Sam E. John, David B. Grayden | Link | Objective. Electroencephalography (EEG) data is derived by sampling continuous neurological time series signals. In order to prepare EEG signals for machine learning, the signal must be divided into manageable segments. The current naive approach uses arbitrary fixed time slices, which may have limited biological relevance because brain states are not confined to fixed intervals. We investigate whether adaptive segmentation methods are beneficial for machine learning EEG analysis. Approach. We introduce a novel adaptive segmentation method, CTXSEG, that creates variable-length segments based on statistical differences in the EEG data and propose ways to use them with modern machine learning approaches that typically require fixed-length input. We assess CTXSEG using controllable synthetic data generated by our novel signal generator CTXGEN. While our CTXSEG method has general utility, we validate it on a real-world use case by applying it to an EEG seizure detection problem. We compare the performance of CTXSEG with fixed-length segmentation in the preprocessing step of a typical EEG machine learning pipeline for seizure detection. Main results. We found that using CTXSEG to prepare EEG data improves seizure detection performance compared to fixed-length approaches when evaluated using a standardized framework, without modifying the machine learning method, and requires fewer segments. Significance. This work demonstrates that adaptive segmentation with CTXSEG can be readily applied to modern machine learning approaches, with potential to improve performance. It is a promising alternative to fixed-length segmentation for signal preprocessing and should be considered as part of the standard preprocessing repertoire in EEG machine learning applications. |
| 2025-08-26 | SecureV2X: An Efficient and Privacy-Preserving System for Vehicle-to-Everything (V2X) Applications | Joshua Lee, Ali Arastehfard, Weiran Liu, Xuegang Ban, Yuan Hong | Link | Autonomous driving and V2X technologies have developed rapidly in the past decade, leading to improved safety and efficiency in modern transportation. These systems interact with extensive networks of vehicles, roadside infrastructure, and cloud resources to support their machine learning capabilities. However, the widespread use of machine learning in V2X systems raises issues over the privacy of the data involved. This is particularly concerning for smart-transit and driver safety applications which can implicitly reveal user locations or explicitly disclose medical data such as EEG signals. To resolve these issues, we propose SecureV2X, a scalable, multi-agent system for secure neural network inferences deployed between the server and each vehicle. Under this setting, we study two multi-agent V2X applications: secure drowsiness detection, and secure red-light violation detection. Our system achieves strong performance relative to baselines, and scales efficiently to support a large number of secure computation interactions simultaneously. For instance, SecureV2X is |
| 2025-08-27 | A Synoptic Review of High-Frequency Oscillations as a Biomarker in Neurodegenerative Disease | Samin Yaser, Mahad Ali, Yang Jiang, VP Nguyen, Jing Xiang, Laura J. Brattain | Link | High Frequency Oscillations (HFOs), rapid bursts of brain activity above 80 Hz, have emerged as a highly specific biomarker for epileptogenic tissue. Recent evidence suggests that HFOs are also present in Alzheimer's Disease (AD), reflecting underlying network hyperexcitability and offering a promising, noninvasive tool for early diagnosis and disease tracking. This synoptic review provides a comprehensive analysis of publicly available electroencephalography (EEG) datasets relevant to HFO research in neurodegenerative disorders. We conducted a bibliometric analysis of 1,222 articles, revealing a significant and growing research interest in HFOs, particularly within the last ten years. We then systematically profile and compare key public datasets, evaluating their participant cohorts, data acquisition parameters, and accessibility, with a specific focus on their technical suitability for HFO analysis. Our comparative synthesis highlights critical methodological heterogeneity across datasets, particularly in sampling frequency and recording paradigms, which poses challenges for cross-study validation, but also offers opportunities for robustness testing. By consolidating disparate information, clarifying nomenclature, and providing a detailed methodological framework, this review serves as a guide for researchers aiming to leverage public data to advance the role of HFOs as a cross-disease biomarker for AD and related conditions. |
| 2025-08-27 | Alljoined-1.6M: A Million-Trial EEG-Image Dataset for Evaluating Affordable Brain-Computer Interfaces | Jonathan Xu, Ugo Bruzadin Nunes, Wangshu Jiang, Samuel Ryther, Jordan Pringle, Paul S. Scotti, Arnaud Delorme, Reese Kneeland | Link | We present a new large-scale electroencephalography (EEG) dataset as part of the THINGS initiative, comprising over 1.6 million visual stimulus trials collected from 20 participants, and totaling more than twice the size of the most popular current benchmark dataset, THINGS-EEG2. Crucially, our data was recorded using a 32-channel consumer-grade wet electrode system costing ~$2.2k, around 27x cheaper than research-grade EEG systems typically used in cognitive neuroscience labs. Our work is one of the first open-source, large-scale EEG resource designed to closely reflect the quality of hardware that is practical to deploy in real-world, downstream applications of brain-computer interfaces (BCIs). We aim to explore the specific question of whether deep neural network-based BCI research and semantic decoding methods can be effectively conducted with such affordable systems, filling an important gap in current literature that is extremely relevant for future research. In our analysis, we not only demonstrate that decoding of high-level semantic information from EEG of visualized images is possible at consumer-grade hardware, but also that our data can facilitate effective EEG-to-Image reconstruction even despite significantly lower signal-to-noise ratios. In addition to traditional benchmarks, we also conduct analyses of EEG-to-Image models that demonstrate log-linear decoding performance with increasing data volume on our data, and discuss the trade-offs between hardware cost, signal fidelity, and the scale of data collection efforts in increasing the size and utility of currently available datasets. Our contributions aim to pave the way for large-scale, cost-effective EEG research with widely accessible equipment, and position our dataset as a unique resource for the democratization and development of effective deep neural models of visual cognition. |
| 2025-08-25 | How do Humans and LLMs Process Confusing Code? | Youssef Abdelsalam, Norman Peitek, Anna-Maria Maurer, Mariya Toneva, Sven Apel | Link | Already today, humans and programming assistants based on large language models (LLMs) collaborate in everyday programming tasks. Clearly, a misalignment between how LLMs and programmers comprehend code can lead to misunderstandings, inefficiencies, low code quality, and bugs. A key question in this space is whether humans and LLMs are confused by the same kind of code. This would not only guide our choices of integrating LLMs in software engineering workflows, but also inform about possible improvements of LLMs. To this end, we conducted an empirical study comparing an LLM to human programmers comprehending clean and confusing code. We operationalized comprehension for the LLM by using LLM perplexity, and for human programmers using neurophysiological responses (in particular, EEG-based fixation-related potentials). We found that LLM perplexity spikes correlate both in terms of location and amplitude with human neurophysiological responses that indicate confusion. This result suggests that LLMs and humans are similarly confused about the code. Based on these findings, we devised a data-driven, LLM-based approach to identify regions of confusion in code that elicit confusion in human programmers. |
| 2025-08-25 | The Effects of Communication Delay on Human Performance and Neurocognitive Responses in Mobile Robot Teleoperation | Zhaokun Chen, Wenshuo Wang, Wenzhuo Liu, Yichen Liu, Junqiang Xi | Link | Communication delays in mobile robot teleoperation adversely affect human-machine collaboration. Understanding delay effects on human operational performance and neurocognition is essential for resolving this issue. However, no previous research has explored this. To fill this gap, we conduct a human-in-the-loop experiment involving 10 participants, integrating electroencephalography (EEG) and robot behavior data under varying delays (0-500 ms in 100 ms increments) to systematically investigate these effects. Behavior analysis reveals significant performance degradation at 200-300 ms delays, affecting both task efficiency and accuracy. EEG analysis discovers features with significant delay dependence: frontal |
| Publish Date | Title | Authors | URL | Abstract |
|---|---|---|---|---|
| 2025-08-29 | A Single Subject Machine Learning Based Classification of Motor Imagery EEGs | Dario Sanalitro, Marco Finocchiaro, Pasquale Memmolo, Emanuela Cutuli, Maide Bucolo | Link | Motor Imagery-Based Brain-Computer Interfaces (MI-BCIs) are systems that detect and interpret brain activity patterns linked to the mental visualization of movement, and then translate these into instructions for controlling external robotic or domotic devices. Such devices have the potential to be useful in a broad variety of applications. While implementing a system that would help individuals restore some freedom levels, the interpretation of (Electroencephalography) EEG data remains a complex and unsolved problem. In the literature, the classification of left and right imagined movements has been extensively studied. This study introduces a novel pipeline that makes use of machine learning techniques for classifying MI EEG data. The entire framework is capable of accurately categorizing left and imagined motions, as well as rest phases, for a set of 52 subjects who performed a MI task. We trained a within subject model on each individual subject. The methodology has been offline evaluated and compared to four studies that are currently the state-of-the-art regarding the specified dataset. The results show that our proposed framework could be used with MI-BCI systems in light of its failsafe classification performances, i.e. 99.5% in accuracy |
| 2025-08-27 | Alljoined-1.6M: A Million-Trial EEG-Image Dataset for Evaluating Affordable Brain-Computer Interfaces | Jonathan Xu, Ugo Bruzadin Nunes, Wangshu Jiang, Samuel Ryther, Jordan Pringle, Paul S. Scotti, Arnaud Delorme, Reese Kneeland | Link | We present a new large-scale electroencephalography (EEG) dataset as part of the THINGS initiative, comprising over 1.6 million visual stimulus trials collected from 20 participants, and totaling more than twice the size of the most popular current benchmark dataset, THINGS-EEG2. Crucially, our data was recorded using a 32-channel consumer-grade wet electrode system costing ~$2.2k, around 27x cheaper than research-grade EEG systems typically used in cognitive neuroscience labs. Our work is one of the first open-source, large-scale EEG resource designed to closely reflect the quality of hardware that is practical to deploy in real-world, downstream applications of brain-computer interfaces (BCIs). We aim to explore the specific question of whether deep neural network-based BCI research and semantic decoding methods can be effectively conducted with such affordable systems, filling an important gap in current literature that is extremely relevant for future research. In our analysis, we not only demonstrate that decoding of high-level semantic information from EEG of visualized images is possible at consumer-grade hardware, but also that our data can facilitate effective EEG-to-Image reconstruction even despite significantly lower signal-to-noise ratios. In addition to traditional benchmarks, we also conduct analyses of EEG-to-Image models that demonstrate log-linear decoding performance with increasing data volume on our data, and discuss the trade-offs between hardware cost, signal fidelity, and the scale of data collection efforts in increasing the size and utility of currently available datasets. Our contributions aim to pave the way for large-scale, cost-effective EEG research with widely accessible equipment, and position our dataset as a unique resource for the democratization and development of effective deep neural models of visual cognition. |
| 2025-08-22 | EEG Study of the Influence of Imagined Temperature Sensations on Neuronal Activity in the Sensorimotor Cortex | Anton Belichenko, Daria Trinitatova, Aigul Nasibullina, Lev Yakovlev, Dzmitry Tsetserukou | Link | Understanding the neural correlates of sensory imagery is crucial for advancing cognitive neuroscience and developing novel Brain-Computer Interface (BCI) paradigms. This study investigated the influence of imagined temperature sensations (ITS) on neural activity within the sensorimotor cortex. The experimental study involved the evaluation of neural activity using electroencephalography (EEG) during both real thermal stimulation (TS: 40{\deg}C Hot, 20{\deg}C Cold) applied to the participants' hand, and the mental temperature imagination (ITS) of the corresponding hot and cold sensations. The analysis focused on quantifying the event-related desynchronization (ERD) of the sensorimotor mu-rhythm (8-13 Hz). The experimental results revealed a characteristic mu-ERD localized over central scalp regions (e.g., C3) during both TS and ITS conditions. Although the magnitude of mu-ERD during ITS was slightly lower than during TS, this difference was not statistically significant (p>.05). However, ERD during both ITS and TS was statistically significantly different from the resting baseline (p<.001). These findings demonstrate that imagining temperature sensations engages sensorimotor cortical mechanisms in a manner comparable to actual thermal perception. This insight expands our understanding of the neurophysiological basis of sensory imagery and suggests the potential utility of ITS for non-motor BCI control and neurorehabilitation technologies. |
| 2025-08-22 | Motor Imagery EEG Signal Classification Using Minimally Random Convolutional Kernel Transform and Hybrid Deep Learning | Jamal Hwaidi, Mohamed Chahine Ghanem | Link | The brain-computer interface (BCI) establishes a non-muscle channel that enables direct communication between the human body and an external device. Electroencephalography (EEG) is a popular non-invasive technique for recording brain signals. It is critical to process and comprehend the hidden patterns linked to a specific cognitive or motor task, for instance, measured through the motor imagery brain-computer interface (MI-BCI). A significant challenge is presented by classifying motor imagery-based electroencephalogram (MI-EEG) tasks, given that EEG signals exhibit nonstationarity, time-variance, and individual diversity. Obtaining good classification accuracy is also very difficult due to the growing number of classes and the natural variability among individuals. To overcome these issues, this paper proposes a novel method for classifying EEG motor imagery signals that extracts features efficiently with Minimally Random Convolutional Kernel Transform (MiniRocket), a linear classifier then uses the extracted features for activity recognition. Furthermore, a novel deep learning based on Convolutional Neural Network (CNN) and Long Short Term Memory (LSTM) architecture to serve as a baseline was proposed and demonstrated that classification via MiniRocket's features achieves higher performance than the best deep learning models at lower computational cost. The PhysioNet dataset was used to evaluate the performance of the proposed approaches. The proposed models achieved mean accuracy values of 98.63% and 98.06% for the MiniRocket and CNN-LSTM, respectively. The findings demonstrate that the proposed approach can significantly enhance motor imagery EEG accuracy and provide new insights into the feature extraction and classification of MI-EEG. |
| 2025-08-20 | Detecting Reading-Induced Confusion Using EEG and Eye Tracking | Haojun Zhuang, Dünya Baradari, Nataliya Kosmyna, Arnav Balyan, Constanze Albrecht, Stephanie Chen, Pattie Maes | Link | Humans regularly navigate an overwhelming amount of information via text media, whether reading articles, browsing social media, or interacting with chatbots. Confusion naturally arises when new information conflicts with or exceeds a reader's comprehension or prior knowledge, posing a challenge for learning. In this study, we present a multimodal investigation of reading-induced confusion using EEG and eye tracking. We collected neural and gaze data from 11 adult participants as they read short paragraphs sampled from diverse, real-world sources. By isolating the N400 event-related potential (ERP), a well-established neural marker of semantic incongruence, and integrating behavioral markers from eye tracking, we provide a detailed analysis of the neural and behavioral correlates of confusion during naturalistic reading. Using machine learning, we show that multimodal (EEG + eye tracking) models improve classification accuracy by 4-22% over unimodal baselines, reaching an average weighted participant accuracy of 77.3% and a best accuracy of 89.6%. Our results highlight the dominance of the brain's temporal regions in these neural signatures of confusion, suggesting avenues for wearable, low-electrode brain-computer interfaces (BCI) for real-time monitoring. These findings lay the foundation for developing adaptive systems that dynamically detect and respond to user confusion, with potential applications in personalized learning, human-computer interaction, and accessibility. |
| 2025-08-18 | Cyber Risks to Next-Gen Brain-Computer Interfaces: Analysis and Recommendations | Tyler Schroder, Renee Sirbu, Sohee Park, Jessica Morley, Sam Street, Luciano Floridi | Link | Brain-computer interfaces (BCIs) show enormous potential for advancing personalized medicine. However, BCIs also introduce new avenues for cyber-attacks or security compromises. In this article, we analyze the problem and make recommendations for device manufacturers to better secure devices and to help regulators understand where more guidance is needed to protect patient safety and data confidentiality. Device manufacturers should implement the prior suggestions in their BCI products. These recommendations help protect BCI users from undue risks, including compromised personal health and genetic information, unintended BCI-mediated movement, and many other cybersecurity breaches. Regulators should mandate non-surgical device update methods, strong authentication and authorization schemes for BCI software modifications, encryption of data moving to and from the brain, and minimize network connectivity where possible. We also design a hypothetical, average-case threat model that identifies possible cybersecurity threats to BCI patients and predicts the likeliness of risk for each category of threat. BCIs are at less risk of physical compromise or attack, but are vulnerable to remote attack; we focus on possible threats via network paths to BCIs and suggest technical controls to limit network connections. |
| 2025-08-16 | Mind the Generation Process: Fine-Grained Confidence Estimation During LLM Generation | Jinyi Han, Tingyun Li, Shisong Chen, Jie Shi, Xinyi Wang, Guanglei Yue, Jiaqing Liang, Xin Lin, Liqian Wen, Zulong Chen, Yanghua Xiao | Link | While large language models (LLMs) have demonstrated remarkable performance across diverse tasks, they fundamentally lack self-awareness and frequently exhibit overconfidence, assigning high confidence scores to incorrect predictions. Accurate confidence estimation is therefore critical for enhancing the trustworthiness and reliability of LLM-generated outputs. However, existing approaches suffer from coarse-grained scoring mechanisms that fail to provide fine-grained, continuous confidence estimates throughout the generation process. To address these limitations, we introduce FineCE, a novel confidence estimation method that delivers accurate, fine-grained confidence scores during text generation. Specifically, we first develop a comprehensive pipeline for constructing training data that effectively captures the underlying probabilistic distribution of LLM responses, and then train a model to predict confidence scores for arbitrary text sequences in a supervised manner. Furthermore, we propose a Backward Confidence Integration (BCI) strategy that leverages information from the subsequent text to enhance confidence estimation for the current sequence during inference. We also introduce three strategies for identifying optimal positions to perform confidence estimation within the generation process. Extensive experiments on multiple benchmark datasets demonstrate that FineCE consistently outperforms existing classical confidence estimation methods. Our code and all baselines used in the paper are available on GitHub. |
| 2025-08-15 | Control of a commercial vehicle by a tetraplegic human using a bimanual brain-computer interface | Xinyun Zou, Jorge Gamez, Meghna Menon, Phillip Ring, Chadwick Boulay, Likhith Chitneni, Jackson Brennecke, Shana R. Melby, Gracy Kureel, Kelsie Pejsa, Emily R. Rosario, Ausaf A. Bari, Aniruddh Ravindran, Tyson Aflalo, Spencer S. Kellis, Dimitar Filev, Florian Solzbacher, Richard A. Andersen | Link | Brain-computer interfaces (BCIs) read neural signals directly from the brain to infer motor planning and execution. However, the implementation of this technology has been largely limited to laboratory settings, with few real-world applications. We developed a bimanual BCI system to drive a vehicle in both simulated and real-world environments. We demonstrate that an individual with tetraplegia, implanted with intracortical BCI electrodes in the posterior parietal cortex (PPC) and the hand knob region of the motor cortex (MC), reacts at least as fast and precisely as motor intact participants, and drives a simulated vehicle as proficiently as the same control group. This BCI participant, living in California, could also remotely drive a Ford Mustang Mach-E vehicle in Michigan. Our first teledriving task relied on cursor control for speed and steering in a closed urban test facility. However, the final BCI system added click control for full-stop braking and thus enabled bimanual cursor-and-click control for both simulated driving through a virtual town with traffic and teledriving through an obstacle course without traffic in the real world. We also demonstrate the safety and feasibility of BCI-controlled driving. This first-of-its-kind implantable BCI application not only highlights the versatility and innovative potentials of BCIs but also illuminates the promising future for the development of life-changing solutions to restore independence to those who suffer catastrophic neurological injury. |
| 2025-08-15 | PTSM: Physiology-aware and Task-invariant Spatio-temporal Modeling for Cross-Subject EEG Decoding | Changhong Jing, Yan Liu, Shuqiang Wang, Bruce X. B. Yu, Gong Chen, Zhejing Hu, Zhi Zhang, Yanyan Shen | Link | Cross-subject electroencephalography (EEG) decoding remains a fundamental challenge in brain-computer interface (BCI) research due to substantial inter-subject variability and the scarcity of subject-invariant representations. This paper proposed PTSM (Physiology-aware and Task-invariant Spatio-temporal Modeling), a novel framework for interpretable and robust EEG decoding across unseen subjects. PTSM employs a dual-branch masking mechanism that independently learns personalized and shared spatio-temporal patterns, enabling the model to preserve individual-specific neural characteristics while extracting task-relevant, population-shared features. The masks are factorized across temporal and spatial dimensions, allowing fine-grained modulation of dynamic EEG patterns with low computational overhead. To further address representational entanglement, PTSM enforces information-theoretic constraints that decompose latent embeddings into orthogonal task-related and subject-related subspaces. The model is trained end-to-end via a multi-objective loss integrating classification, contrastive, and disentanglement objectives. Extensive experiments on cross-subject motor imagery datasets demonstrate that PTSM achieves strong zero-shot generalization, outperforming state-of-the-art baselines without subject-specific calibration. Results highlight the efficacy of disentangled neural representations for achieving both personalized and transferable decoding in non-stationary neurophysiological settings. |
| 2025-08-14 | Codes on any Cayley Graph have an Interactive Oracle Proof of Proximity | Hugo Delavenne, Louise Lallemand | Link | Interactive Oracle Proofs of Proximity (IOPP) are at the heart of code-based SNARKs, a family of zeroknowledge protocols. The first and most famous one is the FRI protocol [BBHR18a], that efficiently tests proximity to Reed-Solomon codes. This paper generalizes the flowering IOPP introduced in [DMR25] for some specific (2, n)-regular Tanner codes to a much broader variety of codes: any code with symbols indexed on the edges of a Cayley graph. The flowering protocol of [DMR25] had a soundness parameter much lower than the FRI protocol [BCI + 23], and complexity parameters that could compete with the FRI [BBHR18a]. The lower soundness and the absence of restriction on the base field may lead to other practical speedups, however the codes considered in [DMR25] have an o(1) minimum distance. The generalization proposed in this paper preserves the soundness parameter with a slight decrease of the complexity parameters, while allowing being applied on codes with constant rate and constant minimum distance thanks to the good expansion properties of some families of Cayley graphs. |
| Publish Date | Title | Authors | URL | Abstract |
|---|---|---|---|---|
| 2025-08-25 | Disentangling the Factors of Convergence between Brains and Computer Vision Models | Joséphine Raugel, Marc Szafraniec, Huy V. Vo, Camille Couprie, Patrick Labatut, Piotr Bojanowski, Valentin Wyart, Jean-Rémi King | Link | Many AI models trained on natural images develop representations that resemble those of the human brain. However, the factors that drive this brain-model similarity remain poorly understood. To disentangle how the model, training and data independently lead a neural network to develop brain-like representations, we trained a family of self-supervised vision transformers (DINOv3) that systematically varied these different factors. We compare their representations of images to those of the human brain recorded with both fMRI and MEG, providing high resolution in spatial and temporal analyses. We assess the brain-model similarity with three complementary metrics focusing on overall representational similarity, topographical organization, and temporal dynamics. We show that all three factors - model size, training amount, and image type - independently and interactively impact each of these brain similarity metrics. In particular, the largest DINOv3 models trained with the most human-centric images reach the highest brain-similarity. This emergence of brain-like representations in AI models follows a specific chronology during training: models first align with the early representations of the sensory cortices, and only align with the late and prefrontal representations of the brain with considerably more training. Finally, this developmental trajectory is indexed by both structural and functional properties of the human cortex: the representations that are acquired last by the models specifically align with the cortical areas with the largest developmental expansion, thickness, least myelination, and slowest timescales. Overall, these findings disentangle the interplay between architecture and experience in shaping how artificial neural networks come to see the world as humans do, thus offering a promising framework to understand how the human brain comes to represent its visual world. |
| 2025-08-20 | The Prompting Brain: Neurocognitive Markers of Expertise in Guiding Large Language Models | Hend Al-Khalifa, Raneem Almansour, Layan Abdulrahman Alhuasini, Alanood Alsaleh, Mohamad-Hani Temsah, Mohamad-Hani_Temsah, Ashwag Rafea S Alruwaili | Link | Prompt engineering has rapidly emerged as a critical skill for effective interaction with large language models (LLMs). However, the cognitive and neural underpinnings of this expertise remain largely unexplored. This paper presents findings from a cross-sectional pilot fMRI study investigating differences in brain functional connectivity and network activity between experts and intermediate prompt engineers. Our results reveal distinct neural signatures associated with higher prompt engineering literacy, including increased functional connectivity in brain regions such as the left middle temporal gyrus and the left frontal pole, as well as altered power-frequency dynamics in key cognitive networks. These findings offer initial insights into the neurobiological basis of prompt engineering proficiency. We discuss the implications of these neurocognitive markers in Natural Language Processing (NLP). Understanding the neural basis of human expertise in interacting with LLMs can inform the design of more intuitive human-AI interfaces, contribute to cognitive models of LLM interaction, and potentially guide the development of AI systems that better align with human cognitive workflows. This interdisciplinary approach aims to bridge the gap between human cognition and machine intelligence, fostering a deeper understanding of how humans learn and adapt to complex AI systems. |
| 2025-08-19 | ASDFormer: A Transformer with Mixtures of Pooling-Classifier Experts for Robust Autism Diagnosis and Biomarker Discovery | Mohammad Izadi, Mehran Safayani | Link | Autism Spectrum Disorder (ASD) is a complex neurodevelopmental condition marked by disruptions in brain connectivity. Functional MRI (fMRI) offers a non-invasive window into large-scale neural dynamics by measuring blood-oxygen-level-dependent (BOLD) signals across the brain. These signals can be modeled as interactions among Regions of Interest (ROIs), which are grouped into functional communities based on their underlying roles in brain function. Emerging evidence suggests that connectivity patterns within and between these communities are particularly sensitive to ASD-related alterations. Effectively capturing these patterns and identifying interactions that deviate from typical development is essential for improving ASD diagnosis and enabling biomarker discovery. In this work, we introduce ASDFormer, a Transformer-based architecture that incorporates a Mixture of Pooling-Classifier Experts (MoE) to capture neural signatures associated with ASD. By integrating multiple specialized expert branches with attention mechanisms, ASDFormer adaptively emphasizes different brain regions and connectivity patterns relevant to autism. This enables both improved classification performance and more interpretable identification of disorder-related biomarkers. Applied to the ABIDE dataset, ASDFormer achieves state-of-the-art diagnostic accuracy and reveals robust insights into functional connectivity disruptions linked to ASD, highlighting its potential as a tool for biomarker discovery. |
| 2025-08-18 | A Dual-Attention Graph Network for fMRI Data Classification | Amirali Arbab, Zeinab Davarani, Mehran Safayani | Link | Understanding the complex neural activity dynamics is crucial for the development of the field of neuroscience. Although current functional MRI classification approaches tend to be based on static functional connectivity or cannot capture spatio-temporal relationships comprehensively, we present a new framework that leverages dynamic graph creation and spatiotemporal attention mechanisms for Autism Spectrum Disorder(ASD) diagnosis. The approach used in this research dynamically infers functional brain connectivity in each time interval using transformer-based attention mechanisms, enabling the model to selectively focus on crucial brain regions and time segments. By constructing time-varying graphs that are then processed with Graph Convolutional Networks (GCNs) and transformers, our method successfully captures both localized interactions and global temporal dependencies. Evaluated on the subset of ABIDE dataset, our model achieves 63.2 accuracy and 60.0 AUC, outperforming static graph-based approaches (e.g., GCN:51.8). This validates the efficacy of joint modeling of dynamic connectivity and spatio-temporal context for fMRI classification. The core novelty arises from (1) attention-driven dynamic graph creation that learns temporal brain region interactions and (2) hierarchical spatio-temporal feature fusion through GCNtransformer fusion. |
| 2025-08-17 | Defining and Benchmarking a Data-Centric Design Space for Brain Graph Construction | Qinwen Ge, Roza G. Bayrak, Anwar Said, Catie Chang, Xenofon Koutsoukos, Tyler Derr | Link | The construction of brain graphs from functional Magnetic Resonance Imaging (fMRI) data plays a crucial role in enabling graph machine learning for neuroimaging. However, current practices often rely on rigid pipelines that overlook critical data-centric choices in how brain graphs are constructed. In this work, we adopt a Data-Centric AI perspective and systematically define and benchmark a data-centric design space for brain graph construction, constrasting with primarily model-centric prior work. We organize this design space into three stages: temporal signal processing, topology extraction, and graph featurization. Our contributions lie less in novel components and more in evaluating how combinations of existing and modified techniques influence downstream performance. Specifically, we study high-amplitude BOLD signal filtering, sparsification and unification strategies for connectivity, alternative correlation metrics, and multi-view node and edge features, such as incorporating lagged dynamics. Experiments on the HCP1200 and ABIDE datasets show that thoughtful data-centric configurations consistently improve classification accuracy over standard pipelines. These findings highlight the critical role of upstream data decisions and underscore the importance of systematically exploring the data-centric design space for graph-based neuroimaging. Our code is available at https://github.com/GeQinwen/DataCentricBrainGraphs. |
| 2025-08-16 | A Wavelet-Based Framework for Mapping Long Memory in Resting-State fMRI: Age-Related Changes in the Hippocampus from the ADHD-200 Dataset | Yasaman Shahhosseini, Cédric Beaulac, Farouk S. Nathoo, Michelle F. Miranda | Link | Functional magnetic resonance imaging (fMRI) time series are known to exhibit long-range temporal dependencies that challenge traditional modeling approaches. In this study, we propose a novel computational pipeline to characterize and interpret these dependencies using a long-memory (LM) framework, which captures the slow, power-law decay of autocorrelation in resting-state fMRI (rs-fMRI) signals. The pipeline involves voxelwise estimation of LM parameters via a wavelet-based Bayesian method, yielding spatial maps that reflect temporal dependence across the brain. These maps are then projected onto a lower-dimensional space via a composite basis and are then related to individual-level covariates through group-level regression. We applied this approach to the ADHD-200 dataset and found significant positive associations between age in children and the LM parameter in the hippocampus, after adjusting for ADHD symptom severity and medication status. These findings complement prior neuroimaging work by linking long-range temporal dependence to developmental changes in memory-related brain regions. Overall, the proposed methodology enables detailed mapping of intrinsic temporal dynamics in rs-fMRI and offers new insights into the relationship between functional signal memory and brain development. |
| 2025-08-15 | Language models align with brain regions that represent concepts across modalities | Maria Ryskina, Greta Tuckute, Alexander Fung, Ashley Malkin, Evelina Fedorenko | Link | Cognitive science and neuroscience have long faced the challenge of disentangling representations of language from representations of conceptual meaning. As the same problem arises in today's language models (LMs), we investigate the relationship between LM--brain alignment and two neural metrics: (1) the level of brain activation during processing of sentences, targeting linguistic processing, and (2) a novel measure of meaning consistency across input modalities, which quantifies how consistently a brain region responds to the same concept across paradigms (sentence, word cloud, image) using an fMRI dataset (Pereira et al., 2018). Our experiments show that both language-only and language-vision models predict the signal better in more meaning-consistent areas of the brain, even when these areas are not strongly sensitive to language processing, suggesting that LMs might internally represent cross-modal conceptual meaning. |
| 2025-08-15 | BRIEF: BRain-Inspired network connection search with Extensive temporal feature Fusion enhances disease classification | Xiangxiang Cui, Min Zhao, Dongmei Zhi, Shile Qi, Vince D Calhoun, Jing Sui | Link | Existing deep learning models for functional MRI-based classification have limitations in network architecture determination (relying on experience) and feature space fusion (mostly simple concatenation, lacking mutual learning). Inspired by the human brain's mechanism of updating neural connections through learning and decision-making, we proposed a novel BRain-Inspired feature Fusion (BRIEF) framework, which is able to optimize network architecture automatically by incorporating an improved neural network connection search (NCS) strategy and a Transformer-based multi-feature fusion module. Specifically, we first extracted 4 types of fMRI temporal representations, i.e., time series (TCs), static/dynamic functional connection (FNC/dFNC), and multi-scale dispersion entropy (MsDE), to construct four encoders. Within each encoder, we employed a modified Q-learning to dynamically optimize the NCS to extract high-level feature vectors, where the NCS is formulated as a Markov Decision Process. Then, all feature vectors were fused via a Transformer, leveraging both stable/time-varying connections and multi-scale dependencies across different brain regions to achieve the final classification. Additionally, an attention module was embedded to improve interpretability. The classification performance of our proposed BRIEF was compared with 21 state-of-the-art models by discriminating two mental disorders from healthy controls: schizophrenia (SZ, n=1100) and autism spectrum disorder (ASD, n=1550). BRIEF demonstrated significant improvements of 2.2% to 12.1% compared to 21 algorithms, reaching an AUC of 91.5% - 0.6% for SZ and 78.4% - 0.5% for ASD, respectively. This is the first attempt to incorporate a brain-inspired, reinforcement learning strategy to optimize fMRI-based mental disorder classification, showing significant potential for identifying precise neuroimaging biomarkers. |
| 2025-08-15 | Estimating Covariate Effects on Functional Connectivity using Voxel-Level fMRI Data | Wei Zhao, Brian J. Reich, Emily C. Hector | Link | Functional connectivity (FC) analysis of resting-state fMRI data provides a framework for characterizing brain networks and their association with participant-level covariates. Due to the high dimensionality of neuroimaging data, standard approaches often average signals within regions of interest (ROIs), which ignores the underlying spatiotemporal dependence among voxels and can lead to biased or inefficient inference. We propose to use a summary statistic -- the empirical voxel-wise correlations between ROIs -- and, crucially, model the complex covariance structure among these correlations through a new positive definite covariance function. Building on this foundation, we develop a computationally efficient two-step estimation procedure that enables statistical inference on covariate effects on region-level connectivity. Simulation studies show calibrated uncertainty quantification, and substantial gains in validity of the statistical inference over the standard averaging method. With data from the Autism Brain Imaging Data Exchange, we show that autism spectrum disorder is associated with altered FC between attention-related ROIs after adjusting for age and gender. The proposed framework offers an interpretable and statistically rigorous approach to estimation of covariate effects on FC suitable for large-scale neuroimaging studies. |
| 2025-08-14 | Insights from the Algonauts 2025 Winners | Paul S. Scotti, Mihir Tripathy | Link | The Algonauts 2025 Challenge just wrapped up a few weeks ago. It is a biennial challenge in computational neuroscience in which teams attempt to build models that predict human brain activity from carefully curated stimuli. Previous editions (2019, 2021, 2023) focused on still images and short videos; the 2025 edition, which concluded last month (late July), pushed the field further by using long, multimodal movies. Teams were tasked with predicting fMRI responses across 1,000 whole-brain parcels across four participants in the dataset who were scanned while watching nearly 80 hours of naturalistic movie stimuli. These recordings came from the CNeuroMod project and included 65 hours of training data, about 55 hours of Friends (seasons 1-6) plus four feature films (The Bourne Supremacy, Hidden Figures, Life, and The Wolf of Wall Street). The remaining data were used for validation: Season 7 of Friends for in-distribution tests, and the final winners for the Challenge were those who could best predict brain activity for six films in their held-out out-of-distribution (OOD) set. The winners were just announced and the top team reports are now publicly available. As members of the MedARC team which placed 4th in the competition, we reflect on the approaches that worked, what they reveal about the current state of brain encoding, and what might come next. |
| Publish Date | Title | Authors | URL | Abstract |
|---|---|---|---|---|
| 2025-08-29 | Quantile Function-Based Models for Neuroimaging Classification Using Wasserstein Regression | Jie Li, Gary Green, Jian Zhang | Link | We propose a novel quantile function-based approach for neuroimaging classification using Wasserstein-Fr\'echet regression, specifically applied to the detection of mild traumatic brain injury (mTBI) based on the MEG and MRI data. Conventional neuroimaging classification methods for mTBI detection typically extract summary statistics from brain signals across the different epochs, which may result in the loss of important distributional information, such as variance, skewness, kurtosis, etc. Our approach treats complete probability density functions of epoch space results as functional response variables within a Wasserstein-Fr\'echet regression framework, thereby preserving the full distributional characteristics of epoch results from |
| 2025-08-25 | Disentangling the Factors of Convergence between Brains and Computer Vision Models | Joséphine Raugel, Marc Szafraniec, Huy V. Vo, Camille Couprie, Patrick Labatut, Piotr Bojanowski, Valentin Wyart, Jean-Rémi King | Link | Many AI models trained on natural images develop representations that resemble those of the human brain. However, the factors that drive this brain-model similarity remain poorly understood. To disentangle how the model, training and data independently lead a neural network to develop brain-like representations, we trained a family of self-supervised vision transformers (DINOv3) that systematically varied these different factors. We compare their representations of images to those of the human brain recorded with both fMRI and MEG, providing high resolution in spatial and temporal analyses. We assess the brain-model similarity with three complementary metrics focusing on overall representational similarity, topographical organization, and temporal dynamics. We show that all three factors - model size, training amount, and image type - independently and interactively impact each of these brain similarity metrics. In particular, the largest DINOv3 models trained with the most human-centric images reach the highest brain-similarity. This emergence of brain-like representations in AI models follows a specific chronology during training: models first align with the early representations of the sensory cortices, and only align with the late and prefrontal representations of the brain with considerably more training. Finally, this developmental trajectory is indexed by both structural and functional properties of the human cortex: the representations that are acquired last by the models specifically align with the cortical areas with the largest developmental expansion, thickness, least myelination, and slowest timescales. Overall, these findings disentangle the interplay between architecture and experience in shaping how artificial neural networks come to see the world as humans do, thus offering a promising framework to understand how the human brain comes to represent its visual world. |
| 2025-08-21 | Probing |
Pratik Adarsh, Rajrupa Banerjee, Purushottam Sahu, Utkarsh Patel, Sudhanwa Patra | Link | Neutrinoless double beta decay ( |
| 2025-08-19 | Reduction of Electromagnetic Interference in ultra-low noise Bimodal MEG & EEG | Jim Barnes, Lukasz Radzinski, Soudabeh Arsalani, Gunnar Waterstraat, Gabriel Curio, Jens Haueisen, Rainer Körber | Link | Single-channel SQUID system technology, operating at a noise level of 100s of aT/$\sqrt{\textrm{Hz}}$, enables the non-invasive detection of synchronized spiking activity at the single-trial level via magnetoencephalography (MEG). However, when combined with simultaneous electroencephalography (EEG) recordings, the noise performance of the ultrasensitive MEG system can be greatly diminished. This issue negates some of the complementary qualities of these two recording methods. In addition, typical electrical components required for electrical stimulation of peripheral nerves, a common method for evoking specific brain responses, are also observed to have a detrimental influence on ultra-low MEG noise performance. These effects are caused by electromagnetic interference (EMI) and typically preclude single-trial detection. This work outlines, how careful design allows a significant reduction of the impact of EMI when these different electronic systems are operated concurrently. This optimization enabled the simultaneous single-trial detection of synchronized spiking activity using these two highly sensitive recording modalities. |
| 2025-07-31 | Wave Turbulence and Cortical Dynamics | Gerald Kaushallye Cooray | Link | Cortical activity recorded through EEG and MEG reflects complex dynamics that span multiple temporal and spatial scales. Spectral analyses of these signals consistently reveal power-law behaviour, a hallmark of turbulent systems. In this paper, we derive a kinetic equation for neural field activity based on wave turbulence theory, highlighting how quantities such as energy and pseudo-particle density flow through wave-space (k-space) via direct and inverse cascades. We explore how different forms of nonlinearity, particularly 3-wave and 4-wave interactions, shape spectral features, including harmonic generation, spectral dispersion, and transient dynamics. While the observed power-law decays in empirical data are broadly consistent with turbulent cascades, variations across studies, such as the presence of dual decay rates or harmonic structures, point to a diversity of underlying mechanisms. We argue that although no single model fully explains all spectral observations, key constraints emerge: namely, that cortical dynamics exhibit features consistent with turbulent wave systems involving both single and dual cascades and a mixture of 3- and 4-wave interactions. This turbulence-based framework offers a principled and unifying approach to interpreting large-scale brain activity, including state transitions and seizure dynamics. |
| 2025-07-29 | Following the Committor Flow: A Data-Driven Discovery of Transition Pathways | Cheng Giuseppe Chen, Chenyu Tang, Alberto Megías, Radu A. Talmazan, Sergio Contreras Arredondo, Benoît Roux, Christophe Chipot | Link | The discovery of transition pathways to unravel distinct reaction mechanisms and, in general, rare events that occur in molecular systems is still a challenge. Recent advances have focused on analyzing the transition path ensemble using the committor probability, widely regarded as the most informative one-dimensional reaction coordinate. Consistency between transition pathways and the committor function is essential for accurate mechanistic insight. In this work, we propose an iterative framework to infer the committor and, subsequently, to identify the most relevant transition pathways. Starting from an initial guess for the transition path, we generate biased sampling from which we train a neural network to approximate the committor probability. From this learned committor, we extract dominant transition channels as discretized strings lying on isocommittor surfaces. These pathways are then used to enhance sampling and iteratively refine both the committor and the transition paths until convergence. The resulting committor enables accurate estimation of the reaction rate constant. We demonstrate the effectiveness of our approach on benchmark systems, including a two-dimensional model potential, peptide conformational transitions, and a Diels--Alder reaction. |
| 2025-07-23 | From Atoms to Dynamics: Learning the Committor Without Collective Variables | Sergio Contreras Arredondo, Chenyu Tang, Radu A. Talmazan, Alberto Megías, Cheng Giuseppe Chen, Christophe Chipot | Link | This Brief Communication introduces a graph-neural-network architecture built on geometric vector perceptrons to predict the committor function directly from atomic coordinates, bypassing the need for hand-crafted collective variables (CVs). The method offers atom-level interpretability, pinpointing the key atomic players in complex transitions without relying on prior assumptions. Applied across diverse molecular systems, the method accurately infers the committor function and highlights the importance of each heavy atom in the transition mechanism. It also yields precise estimates of the rate constants for the underlying processes. The proposed approach opens new avenues for understanding and modeling complex dynamics, by enabling CV-free learning and automated identification of physically meaningful reaction coordinates of complex molecular processes. |
| 2025-07-16 | Diffusion-based translation between unpaired spontaneous premature neonatal EEG and fetal MEG | Benoît Brebion, Alban Gallard, Katrin Sippel, Amer Zaylaa, Hubert Preissl, Sahar Moghimi, Fabrice Wallois, Yaël Frégier | Link | Background and objective: Brain activity in premature newborns has traditionally been studied using electroencephalography (EEG), leading to substantial advances in our understanding of early neural development. However, since brain development takes root at the fetal stage, a critical window of this process remains largely unknown. The only technique capable of recording neural activity in the intrauterine environment is fetal magnetoencephalography (fMEG), but this approach presents challenges in terms of data quality and scarcity. Using artificial intelligence, the present research aims to transfer the well-established knowledge from EEG studies to fMEG to improve understanding of prenatal brain development, laying the foundations for better detection and treatment of potential pathologies. Methods: We developed an unpaired diffusion translation method based on dual diffusion bridges, which notably includes numerical integration improvements to obtain more qualitative results at a lower computational cost. Models were trained on our unpaired dataset of bursts of spontaneous activity from 30 high-resolution premature newborns EEG recordings and 44 fMEG recordings. Results: We demonstrate that our method achieves significant improvement upon previous results obtained with Generative Adversarial Networks (GANs), by almost 5% on the mean squared error in the time domain, and completely eliminating the mode collapse problem in the frequency domain, thus achieving near-perfect signal fidelity. Conclusion: We set a new state of the art in the EEG-fMEG unpaired translation problem, as our developed tool completely paves the way for early brain activity analysis. Overall, we also believe that our method could be reused for other unpaired signal translation applications. |
| 2025-07-13 | BrainFLORA: Uncovering Brain Concept Representation via Multimodal Neural Embeddings | Dongyang Li, Haoyang Qin, Mingyang Wu, Chen Wei, Quanying Liu | Link | Understanding how the brain represents visual information is a fundamental challenge in neuroscience and artificial intelligence. While AI-driven decoding of neural data has provided insights into the human visual system, integrating multimodal neuroimaging signals, such as EEG, MEG, and fMRI, remains a critical hurdle due to their inherent spatiotemporal misalignment. Current approaches often analyze these modalities in isolation, limiting a holistic view of neural representation. In this study, we introduce BrainFLORA, a unified framework for integrating cross-modal neuroimaging data to construct a shared neural representation. Our approach leverages multimodal large language models (MLLMs) augmented with modality-specific adapters and task decoders, achieving state-of-the-art performance in joint-subject visual retrieval task and has the potential to extend multitasking. Combining neuroimaging analysis methods, we further reveal how visual concept representations align across neural modalities and with real world object perception. We demonstrate that the brain's structured visual concept representations exhibit an implicit mapping to physical-world stimuli, bridging neuroscience and machine learning from different modalities of neural imaging. Beyond methodological advancements, BrainFLORA offers novel implications for cognitive neuroscience and brain-computer interfaces (BCIs). Our code is available at https://github.com/ncclab-sustech/BrainFLORA. |
| 2025-07-09 | Resonant leptogenesis in inverse see-saw framework with modular |
Abhishek, V. Suryanarayana Mummidi | Link | This work introduces a model for lepton mass generation and flavor mixing, realized through a (2,3) inverse seesaw structure within a modular \( S_4 \) symmetry framework. The model employs modular forms to construct the lepton Yukawa couplings, thereby significantly simplifying the model by reducing its complexity. A detailed numerical analysis demonstrates consistency with current neutrino oscillation data, yielding constrained predictions for the mixing angles and CP-violating phases. The Dirac CP phase is sharply localized near \( \delta_{\rm CP} \sim 359^\circ \), and the model predicts an effective Majorana mass \( |m_{ee}| \sim \mathcal{O}(10^{-3}) \,\text{eV} \), Within the scope of upcoming experiments on neutrinoless double beta decay such as nEXO and AMoRE-II. The model also remains consistent with current bounds on charged lepton flavor violating processes from MEG and BaBar. We further explore resonant leptogenesis enabled by quasi-degenerate heavy neutrino states, and show that observed baryon asymmetry of the universe can be succesfully generated within this framework. The combined treatment of low-energy observables and high-scale baryogenesis demonstrates the predictivity and testability of the modular \( S_4 \)-based ISS(2,3) framework. |
| Publish Date | Title | Authors | URL | Abstract |
|---|---|---|---|---|
| 2025-07-09 | Quantifying Uncertainty in Error Consistency: Towards Reliable Behavioral Comparison of Classifiers | Thomas Klein, Sascha Meyen, Wieland Brendel, Felix A. Wichmann, Kristof Meding | Link | Benchmarking models is a key factor for the rapid progress in machine learning (ML) research. Thus, further progress depends on improving benchmarking metrics. A standard metric to measure the behavioral alignment between ML models and human observers is error consistency (EC). EC allows for more fine-grained comparisons of behavior than other metrics such as e.g. accuracy, and has been used in the influential Brain-Score benchmark to rank different DNNs by their behavioral consistency with humans. Previously, EC values have been reported without confidence intervals. However, empirically measured EC values are typically noisy -- thus, without confidence intervals, valid benchmarking conclusions are problematic. Here we improve on standard EC in two ways: First, we show how to obtain confidence intervals for EC using a bootstrapping technique, allowing us to derive significance tests for EC. Second, we propose a new computational model relating the EC between two classifiers to the implicit probability that one of them copies responses from the other. This view of EC allows us to give practical guidance to scientists regarding the number of trials required for sufficiently powerful, conclusive experiments. Finally, we use our methodology to revisit popular NeuroAI-results. We find that while the general trend of behavioral differences between humans and machines holds up to scrutiny, many reported differences between deep vision models are statistically insignificant. Our methodology enables researchers to design adequately powered experiments that can reliably detect behavioral differences between models, providing a foundation for more rigorous benchmarking of behavioral alignment. |
| 2025-07-02 | What Neuroscience Can Teach AI About Learning in Continuously Changing Environments | Daniel Durstewitz, Bruno Averbeck, Georgia Koppe | Link | Modern AI models, such as large language models, are usually trained once on a huge corpus of data, potentially fine-tuned for a specific task, and then deployed with fixed parameters. Their training is costly, slow, and gradual, requiring billions of repetitions. In stark contrast, animals continuously adapt to the ever-changing contingencies in their environments. This is particularly important for social species, where behavioral policies and reward outcomes may frequently change in interaction with peers. The underlying computational processes are often marked by rapid shifts in an animal's behaviour and rather sudden transitions in neuronal population activity. Such computational capacities are of growing importance for AI systems operating in the real world, like those guiding robots or autonomous vehicles, or for agentic AI interacting with humans online. Can AI learn from neuroscience? This Perspective explores this question, integrating the literature on continual and in-context learning in AI with the neuroscience of learning on behavioral tasks with shifting rules, reward probabilities, or outcomes. We will outline an agenda for how specifically insights from neuroscience may inform current developments in AI in this area, and - vice versa - what neuroscience may learn from AI, contributing to the evolving field of NeuroAI. |
| 2025-06-12 | NOBLE -- Neural Operator with Biologically-informed Latent Embeddings to Capture Experimental Variability in Biological Neuron Models | Luca Ghafourpour, Valentin Duruisseaux, Bahareh Tolooshams, Philip H. Wong, Costas A. Anastassiou, Anima Anandkumar | Link | Characterizing the diverse computational properties of human neurons via multimodal electrophysiological, transcriptomic, and morphological data provides the foundation for constructing and validating bio-realistic neuron models that can advance our understanding of fundamental mechanisms underlying brain function. However, current modeling approaches remain constrained by the limited availability and intrinsic variability of experimental neuronal data. To capture variability, ensembles of deterministic models are often used, but are difficult to scale as model generation requires repeating computationally expensive optimization for each neuron. While deep learning is becoming increasingly relevant in this space, it fails to capture the full biophysical complexity of neurons, their nonlinear voltage dynamics, and variability. To address these shortcomings, we introduce NOBLE, a neural operator framework that learns a mapping from a continuous frequency-modulated embedding of interpretable neuron features to the somatic voltage response induced by current injection. Trained on data generated from biophysically realistic neuron models, NOBLE predicts distributions of neural dynamics accounting for the intrinsic experimental variability. Unlike conventional bio-realistic neuron models, interpolating within the embedding space offers models whose dynamics are consistent with experimentally observed responses. NOBLE is the first scaled-up deep learning framework validated on real experimental data, enabling efficient generation of synthetic neurons that exhibit trial-to-trial variability and achieve a |
| 2025-05-21 | SynEVO: A neuro-inspired spatiotemporal evolutional framework for cross-domain adaptation | Jiayue Liu, Zhongchao Yi, Zhengyang Zhou, Qihe Huang, Kuo Yang, Xu Wang, Yang Wang | Link | Discovering regularities from spatiotemporal systems can benefit various scientific and social planning. Current spatiotemporal learners usually train an independent model from a specific source data that leads to limited transferability among sources, where even correlated tasks requires new design and training. The key towards increasing cross-domain knowledge is to enable collective intelligence and model evolution. In this paper, inspired by neuroscience theories, we theoretically derive the increased information boundary via learning cross-domain collective intelligence and propose a Synaptic EVOlutional spatiotemporal network, SynEVO, where SynEVO breaks the model independence and enables cross-domain knowledge to be shared and aggregated. Specifically, we first re-order the sample groups to imitate the human curriculum learning, and devise two complementary learners, elastic common container and task-independent extractor to allow model growth and task-wise commonality and personality disentanglement. Then an adaptive dynamic coupler with a new difference metric determines whether the new sample group should be incorporated into common container to achieve model evolution under various domains. Experiments show that SynEVO improves the generalization capacity by at most 42% under cross-domain scenarios and SynEVO provides a paradigm of NeuroAI for knowledge transfer and adaptation. |
| 2025-03-08 | A 7T fMRI dataset of synthetic images for out-of-distribution modeling of vision | Alessandro T. Gifford, Radoslaw M. Cichy, Thomas Naselaris, Kendrick Kay | Link | Large-scale visual neural datasets such as the Natural Scenes Dataset (NSD) are boosting NeuroAI research by enabling computational models of the brain with performances beyond what was possible just a decade ago. However, these datasets lack out-of-distribution (OOD) components, which are crucial for the development of more robust models. Here, we address this limitation by releasing NSD-synthetic, a dataset consisting of 7T fMRI responses from the eight NSD subjects for 284 carefully controlled synthetic images. We show that NSD-synthetic's fMRI responses reliably encode stimulus-related information and are OOD with respect to NSD. Furthermore, OOD generalization tests on NSD-synthetic reveal differences between models of the brain that are not detected with NSD - specifically, self-supervised deep neural networks better explain neural responses than their task-supervised counterparts. These results showcase how NSD-synthetic enables OOD generalization tests that facilitate the development of more robust models of visual processing, and the formulation of more accurate theories of human vision. |
| 2025-02-22 | Brain-Model Evaluations Need the NeuroAI Turing Test | Jenelle Feather, Meenakshi Khosla, N. Apurva Ratan Murty, Aran Nayebi | Link | What makes an artificial system a good model of intelligence? The classical test proposed by Alan Turing focuses on behavior, requiring that an artificial agent's behavior be indistinguishable from that of a human. While behavioral similarity provides a strong starting point, two systems with very different internal representations can produce the same outputs. Thus, in modeling biological intelligence, the field of NeuroAI often aims to go beyond behavioral similarity and achieve representational convergence between a model's activations and the measured activity of a biological system. This position paper argues that the standard definition of the Turing Test is incomplete for NeuroAI, and proposes a stronger framework called the ``NeuroAI Turing Test'', a benchmark that extends beyond behavior alone and \emph{additionally} requires models to produce internal neural representations that are empirically indistinguishable from those of a brain up to measured individual variability, i.e. the differences between a computational model and the brain is no more than the difference between one brain and another brain. While the brain is not necessarily the ceiling of intelligence, it remains the only universally agreed-upon example, making it a natural reference point for evaluating computational models. By proposing this framework, we aim to shift the discourse from loosely defined notions of brain inspiration to a systematic and testable standard centered on both behavior and internal representations, providing a clear benchmark for neuroscientific modeling and AI development. |
| 2025-01-04 | Asynchronous Hebbian/anti-Hebbian networks | Henrique Reis Aguiar, Matthias H. Hennig | Link | Lateral inhibition models coupled with Hebbian plasticity have been shown to learn factorised causal representations of input stimuli, for instance, oriented edges are learned from natural images. Currently, these models require the recurrent dynamics to settle into a stable state before weight changes can be applied, which is not only biologically implausible, but also impractical for real-time learning systems. Here, we propose a new Hebbian learning rule which is implemented using plausible biological mechanisms that have been observed experimentally. We find that this rule allows for efficient, time-continuous learning of factorised representations, very similar to the classic noncontinuous Hebbian/anti-Hebbian learning. Furthermore, we show that this rule naturally prevents catastrophic forgetting when stimuli from different distributions are shown sequentially. |
| 2025-04-03 | NeuroAI for AI Safety | Patrick Mineault, Niccolò Zanichelli, Joanne Zichen Peng, Anton Arkhipov, Eli Bingham, Julian Jara-Ettinger, Emily Mackevicius, Adam Marblestone, Marcelo Mattar, Andrew Payne, Sophia Sanborn, Karen Schroeder, Zenna Tavares, Andreas Tolias, Anthony Zador | Link | As AI systems become increasingly powerful, the need for safe AI has become more pressing. Humans are an attractive model for AI safety: as the only known agents capable of general intelligence, they perform robustly even under conditions that deviate significantly from prior experiences, explore the world safely, understand pragmatics, and can cooperate to meet their intrinsic goals. Intelligence, when coupled with cooperation and safety mechanisms, can drive sustained progress and well-being. These properties are a function of the architecture of the brain and the learning algorithms it implements. Neuroscience may thus hold important keys to technical AI safety that are currently underexplored and underutilized. In this roadmap, we highlight and critically evaluate several paths toward AI safety inspired by neuroscience: emulating the brain's representations, information processing, and architecture; building robust sensory and motor systems from imitating brain data and bodies; fine-tuning AI systems on brain data; advancing interpretability using neuroscience methods; and scaling up cognitively-inspired architectures. We make several concrete recommendations for how neuroscience can positively impact AI safety. |
| 2024-11-21 | Evaluating Representational Similarity Measures from the Lens of Functional Correspondence | Yiqing Bo, Ansh Soni, Sudhanshu Srivastava, Meenakshi Khosla | Link | Neuroscience and artificial intelligence (AI) both face the challenge of interpreting high-dimensional neural data, where the comparative analysis of such data is crucial for revealing shared mechanisms and differences between these complex systems. Despite the widespread use of representational comparisons and the abundance classes of comparison methods, a critical question remains: which metrics are most suitable for these comparisons? While some studies evaluate metrics based on their ability to differentiate models of different origins or constructions (e.g., various architectures), another approach is to assess how well they distinguish models that exhibit distinct behaviors. To investigate this, we examine the degree of alignment between various representational similarity measures and behavioral outcomes, employing group statistics and a comprehensive suite of behavioral metrics for comparison. In our evaluation of eight commonly used representational similarity metrics in the visual domain -- spanning alignment-based, Canonical Correlation Analysis (CCA)-based, inner product kernel-based, and nearest-neighbor methods -- we found that metrics like linear Centered Kernel Alignment (CKA) and Procrustes distance, which emphasize the overall geometric structure or shape of representations, excelled in differentiating trained from untrained models and aligning with behavioral measures, whereas metrics such as linear predictivity, commonly used in neuroscience, demonstrated only moderate alignment with behavior. These insights are crucial for selecting metrics that emphasize behaviorally meaningful comparisons in NeuroAI research. |
| 2024-09-09 | Evidence from fMRI Supports a Two-Phase Abstraction Process in Language Models | Emily Cheng, Richard J. Antonello | Link | Research has repeatedly demonstrated that intermediate hidden states extracted from large language models are able to predict measured brain response to natural language stimuli. Yet, very little is known about the representation properties that enable this high prediction performance. Why is it the intermediate layers, and not the output layers, that are most capable for this unique and highly general transfer task? In this work, we show that evidence from language encoding models in fMRI supports the existence of a two-phase abstraction process within LLMs. We use manifold learning methods to show that this abstraction process naturally arises over the course of training a language model and that the first "composition" phase of this abstraction process is compressed into fewer layers as training continues. Finally, we demonstrate a strong correspondence between layerwise encoding performance and the intrinsic dimensionality of representations from LLMs. We give initial evidence that this correspondence primarily derives from the inherent compositionality of LLMs and not their next-word prediction properties. |
| Publish Date | Title | Authors | URL | Abstract |
|---|---|---|---|---|
| 2025-08-29 | TMUAD: Enhancing Logical Capabilities in Unified Anomaly Detection Models with a Text Memory Bank | Jiawei Liu, Jiahe Hou, Wei Wang, Jinsong Du, Yang Cong, Huijie Fan | Link | Anomaly detection, which aims to identify anomalies deviating from normal patterns, is challenging due to the limited amount of normal data available. Unlike most existing unified methods that rely on carefully designed image feature extractors and memory banks to capture logical relationships between objects, we introduce a text memory bank to enhance the detection of logical anomalies. Specifically, we propose a Three-Memory framework for Unified structural and logical Anomaly Detection (TMUAD). First, we build a class-level text memory bank for logical anomaly detection by the proposed logic-aware text extractor, which can capture rich logical descriptions of objects from input images. Second, we construct an object-level image memory bank that preserves complete object contours by extracting features from segmented objects. Third, we employ visual encoders to extract patch-level image features for constructing a patch-level memory bank for structural anomaly detection. These three complementary memory banks are used to retrieve and compare normal images that are most similar to the query image, compute anomaly scores at multiple levels, and fuse them into a final anomaly score. By unifying structural and logical anomaly detection through collaborative memory banks, TMUAD achieves state-of-the-art performance across seven publicly available datasets involving industrial and medical domains. The model and code are available at https://github.com/SIA-IDE/TMUAD. |
| 2025-08-29 | MoE-Health: A Mixture of Experts Framework for Robust Multimodal Healthcare Prediction | Xiaoyang Wang, Christopher C. Yang | Link | Healthcare systems generate diverse multimodal data, including Electronic Health Records (EHR), clinical notes, and medical images. Effectively leveraging this data for clinical prediction is challenging, particularly as real-world samples often present with varied or incomplete modalities. Existing approaches typically require complete modality data or rely on manual selection strategies, limiting their applicability in real-world clinical settings where data availability varies across patients and institutions. To address these limitations, we propose MoE-Health, a novel Mixture of Experts framework designed for robust multimodal fusion in healthcare prediction. MoE-Health architecture is specifically developed to handle samples with differing modalities and improve performance on critical clinical tasks. By leveraging specialized expert networks and a dynamic gating mechanism, our approach dynamically selects and combines relevant experts based on available data modalities, enabling flexible adaptation to varying data availability scenarios. We evaluate MoE-Health on the MIMIC-IV dataset across three critical clinical prediction tasks: in-hospital mortality prediction, long length of stay, and hospital readmission prediction. Experimental results demonstrate that MoE-Health achieves superior performance compared to existing multimodal fusion methods while maintaining robustness across different modality availability patterns. The framework effectively integrates multimodal information, offering improved predictive performance and robustness in handling heterogeneous and incomplete healthcare data, making it particularly suitable for deployment in diverse healthcare environments with heterogeneous data availability. |
| 2025-08-29 | A Multi-Stage Fine-Tuning and Ensembling Strategy for Pancreatic Tumor Segmentation in Diagnostic and Therapeutic MRI | Omer Faruk Durugol, Maximilian Rokuss, Yannick Kirchhoff, Klaus H. Maier-Hein | Link | Automated segmentation of Pancreatic Ductal Adenocarcinoma (PDAC) from MRI is critical for clinical workflows but is hindered by poor tumor-tissue contrast and a scarcity of annotated data. This paper details our submission to the PANTHER challenge, addressing both diagnostic T1-weighted (Task 1) and therapeutic T2-weighted (Task 2) segmentation. Our approach is built upon the nnU-Net framework and leverages a deep, multi-stage cascaded pre-training strategy, starting from a general anatomical foundation model and sequentially fine-tuning on CT pancreatic lesion datasets and the target MRI modalities. Through extensive five-fold cross-validation, we systematically evaluated data augmentation schemes and training schedules. Our analysis revealed a critical trade-off, where aggressive data augmentation produced the highest volumetric accuracy, while default augmentations yielded superior boundary precision (achieving a state-of-the-art MASD of 5.46 mm and HD95 of 17.33 mm for Task 1). For our final submission, we exploited this finding by constructing custom, heterogeneous ensembles of specialist models, essentially creating a mix of experts. This metric-aware ensembling strategy proved highly effective, achieving a top cross-validation Tumor Dice score of 0.661 for Task 1 and 0.523 for Task 2. Our work presents a robust methodology for developing specialized, high-performance models in the context of limited data and complex medical imaging tasks (Team MIC-DKFZ). |
| 2025-08-29 | Sequential Fair Allocation With Replenishments: A Little Envy Goes An Exponentially Long Way | Chido Onyeze, Sean R. Sinclair, Chamsi Hssaine, Siddhartha Banerjee | Link | We study the trade-off between envy and inefficiency in repeated resource allocation settings with stochastic replenishments, motivated by real-world systems such as food banks and medical supply chains. Specifically, we consider a model in which a decision-maker faced with stochastic demand and resource donations must trade off between an equitable and efficient allocation of resources over an infinite horizon. The decision-maker has access to storage with fixed capacity |
| 2025-08-29 | Neural Network Acceleration on MPSoC board: Integrating SLAC's SNL, Rogue Software and Auto-SNL | Hamza Ezzaoui Rahali, Abhilasha Dave, Larry Ruckman, Mohammad Mehdi Rahimifar, Audrey C. Therrien, James J. Russel, Ryan T. Herbst | Link | The LCLS-II Free Electron Laser (FEL) will generate X-ray pulses for beamline experiments at rates of up to 1~MHz, with detectors producing data throughputs exceeding 1 TB/s. Managing such massive data streams presents significant challenges, as transmission and storage infrastructures become prohibitively expensive. Machine learning (ML) offers a promising solution for real-time data reduction, but conventional implementations introduce excessive latency, making them unsuitable for high-speed experimental environments. To address these challenges, SLAC developed the SLAC Neural Network Library (SNL), a specialized framework designed to deploy real-time ML inference models on Field-Programmable Gate Arrays (FPGA). SNL's key feature is the ability to dynamically update model weights without requiring FPGA resynthesis, enhancing flexibility for adaptive learning applications. To further enhance usability and accessibility, we introduce Auto-SNL, a Python extension that streamlines the process of converting Python-based neural network models into SNL-compatible high-level synthesis code. This paper presents a benchmark comparison against hls4ml, the current state-of-the-art tool, across multiple neural network architectures, fixed-point precisions, and synthesis configurations targeting a Xilinx ZCU102 FPGA. The results showed that SNL achieves competitive or superior latency in most tested architectures, while in some cases also offering FPGA resource savings. This adaptation demonstrates SNL's versatility, opening new opportunities for researchers and academics in fields such as high-energy physics, medical imaging, robotics, and many more. |
| 2025-08-29 | Developer Insights into Designing AI-Based Computer Perception Tools | Maya Guhan, Meghan E. Hurley, Eric A. Storch, John Herrington, Casey Zampella, Julia Parish-Morris, Gabriel Lázaro-Muñoz, Kristin Kostick-Quenet | Link | Artificial intelligence (AI)-based computer perception (CP) technologies use mobile sensors to collect behavioral and physiological data for clinical decision-making. These tools can reshape how clinical knowledge is generated and interpreted. However, effective integration of these tools into clinical workflows depends on how developers balance clinical utility with user acceptability and trustworthiness. Our study presents findings from 20 in-depth interviews with developers of AI-based CP tools. Interviews were transcribed and inductive, thematic analysis was performed to identify 4 key design priorities: 1) to account for context and ensure explainability for both patients and clinicians; 2) align tools with existing clinical workflows; 3) appropriately customize to relevant stakeholders for usability and acceptability; and 4) push the boundaries of innovation while aligning with established paradigms. Our findings highlight that developers view themselves as not merely technical architects but also ethical stewards, designing tools that are both acceptable by users and epistemically responsible (prioritizing objectivity and pushing clinical knowledge forward). We offer the following suggestions to help achieve this balance: documenting how design choices around customization are made, defining limits for customization choices, transparently conveying information about outputs, and investing in user training. Achieving these goals will require interdisciplinary collaboration between developers, clinicians, and ethicists. |
| 2025-08-29 | Leveraging Imperfection with MEDLEY A Multi-Model Approach Harnessing Bias in Medical AI | Farhad Abtahi, Mehdi Astaraki, Fernando Seoane | Link | Bias in medical artificial intelligence is conventionally viewed as a defect requiring elimination. However, human reasoning inherently incorporates biases shaped by education, culture, and experience, suggesting their presence may be inevitable and potentially valuable. We propose MEDLEY (Medical Ensemble Diagnostic system with Leveraged diversitY), a conceptual framework that orchestrates multiple AI models while preserving their diverse outputs rather than collapsing them into a consensus. Unlike traditional approaches that suppress disagreement, MEDLEY documents model-specific biases as potential strengths and treats hallucinations as provisional hypotheses for clinician verification. A proof-of-concept demonstrator was developed using over 30 large language models, creating a minimum viable product that preserved both consensus and minority views in synthetic cases, making diagnostic uncertainty and latent biases transparent for clinical oversight. While not yet a validated clinical tool, the demonstration illustrates how structured diversity can enhance medical reasoning under clinician supervision. By reframing AI imperfection as a resource, MEDLEY offers a paradigm shift that opens new regulatory, ethical, and innovation pathways for developing trustworthy medical AI systems. |
| 2025-08-29 | Temporal Flow Matching for Learning Spatio-Temporal Trajectories in 4D Longitudinal Medical Imaging | Nico Albert Disch, Yannick Kirchhoff, Robin Peretzke, Maximilian Rokuss, Saikat Roy, Constantin Ulrich, David Zimmerer, Klaus Maier-Hein | Link | Understanding temporal dynamics in medical imaging is crucial for applications such as disease progression modeling, treatment planning and anatomical development tracking. However, most deep learning methods either consider only single temporal contexts, or focus on tasks like classification or regression, limiting their ability for fine-grained spatial predictions. While some approaches have been explored, they are often limited to single timepoints, specific diseases or have other technical restrictions. To address this fundamental gap, we introduce Temporal Flow Matching (TFM), a unified generative trajectory method that (i) aims to learn the underlying temporal distribution, (ii) by design can fall back to a nearest image predictor, i.e. predicting the last context image (LCI), as a special case, and (iii) supports |
| 2025-08-29 | MedShift: Implicit Conditional Transport for X-Ray Domain Adaptation | Francisco Caetano, Christiaan Viviers, Peter H. H. de With, Fons van der Sommen | Link | Synthetic medical data offers a scalable solution for training robust models, but significant domain gaps limit its generalizability to real-world clinical settings. This paper addresses the challenge of cross-domain translation between synthetic and real X-ray images of the head, focusing on bridging discrepancies in attenuation behavior, noise characteristics, and soft tissue representation. We propose MedShift, a unified class-conditional generative model based on Flow Matching and Schrodinger Bridges, which enables high-fidelity, unpaired image translation across multiple domains. Unlike prior approaches that require domain-specific training or rely on paired data, MedShift learns a shared domain-agnostic latent space and supports seamless translation between any pair of domains seen during training. We introduce X-DigiSkull, a new dataset comprising aligned synthetic and real skull X-rays under varying radiation doses, to benchmark domain translation models. Experimental results demonstrate that, despite its smaller model size compared to diffusion-based approaches, MedShift offers strong performance and remains flexible at inference time, as it can be tuned to prioritize either perceptual fidelity or structural consistency, making it a scalable and generalizable solution for domain adaptation in medical imaging. The code and dataset are available at https://caetas.github.io/medshift.html |
| 2025-08-29 | Med-RewardBench: Benchmarking Reward Models and Judges for Medical Multimodal Large Language Models | Meidan Ding, Jipeng Zhang, Wenxuan Wang, Cheng-Yi Li, Wei-Chieh Fang, Hsin-Yu Wu, Haiqin Zhong, Wenting Chen, Linlin Shen | Link | Multimodal large language models (MLLMs) hold significant potential in medical applications, including disease diagnosis and clinical decision-making. However, these tasks require highly accurate, context-sensitive, and professionally aligned responses, making reliable reward models and judges critical. Despite their importance, medical reward models (MRMs) and judges remain underexplored, with no dedicated benchmarks addressing clinical requirements. Existing benchmarks focus on general MLLM capabilities or evaluate models as solvers, neglecting essential evaluation dimensions like diagnostic accuracy and clinical relevance. To address this, we introduce Med-RewardBench, the first benchmark specifically designed to evaluate MRMs and judges in medical scenarios. Med-RewardBench features a multimodal dataset spanning 13 organ systems and 8 clinical departments, with 1,026 expert-annotated cases. A rigorous three-step process ensures high-quality evaluation data across six clinically critical dimensions. We evaluate 32 state-of-the-art MLLMs, including open-source, proprietary, and medical-specific models, revealing substantial challenges in aligning outputs with expert judgment. Additionally, we develop baseline models that demonstrate substantial performance improvements through fine-tuning. |