The Medical MNIST is a dataset comprised of 2D and 3D biomedical images used for the purpose of classification. The model.ipynb file shows the use of a custom-made convolutional neural network (CNN) using residual blocks to classify 6 parts of the body from x-ray images, ultimately achieving 99% accuracy. Below is an example of the categorization done by the CNN of the images.
