 This research proposes a novel method using a Kinect V.2 camera and machine learning algorithms to detect Alzheimer's disease, AD. The proposed system was evaluated on a data set of 47 healthy controls, HC, and 38 AD patients. The system was able to accurately distinguish between the two groups with an average accuracy of 97.75% and an F score of 97.67%. This demonstrates the potential of this system as a low-cost and convenient AD assessment tool that can be used during routine checkups or even at home. Future studies should investigate the clinical diagnostic value of this system in a larger cohort. This article was authored by Mahmood Siafalahai, Afsun Hassani Marabin, James E. Galvin and others.