Automated Alzheimer's Disease Classification
Brain Magnetic Resonance Imaging (MRIs) conducted for any reason should be processed through automated Alzheimer's Disease (AD) machine learning algorithms. A very basic ResNet-50 was able to classify 98.99% of all AD cases accurately. Without imagery, extreme gradient boosting classified 91.3% correctly based largely on the Mini-Mental State Exam. (Kudos to the inventors of that scale!) In the coming weeks, I will post many, many other cases where providcrs should reconsider their roles, from histopathology (e.g., breast cancer border screening) to intestinal cancer to..... ML: it is what's for dinner. Are you hungry yet?