Alzheimer’s affects roughly 70% of people with dementia, but it can be difficult and timely to diagnose. Current methods for diagnosis depend on cognitive tests and scans, but they aren’t very accurate and can take a long time, leaving the disease to progress without treatment. A recent Medical News Today article discussed a new algorithm, designed at Imperial College London, that when paired with a single brain MRI scan, can rapidly detect early signs of Alzheimer’s.
The researchers based the algorithm on ones used for classifying cancer tumors. They divided the brain into 115 regions, and designated 660 features like shape, size and texture to each of the regions. They then trained the algorithm to predict Alzheimer’s disease by noting changes in the features from a standard MRI scan. The method was tested on more than 400 patients, and accurately predicted the disease in 98% of cases, and differentiated between early and late-stage Alzheimer’s 79% of the time.