Use of Machine Learning algorithms in Medical Image Processing is challenging and innovative research oriented prject.
As a part of the project, I tried using pre-trained CNN models including:
Feature Extraction using Image Processing
keras, sklearn, pandas, numpy, PyTorch OPenCV
These pre-trained models gave results which suffered from overfitting. Further feature extraction and image processing followed by GANs/ encoder may improve the accuracy.
The dataset was collected from OASIS https://www.oasis-brains.org/