Machine learning and artificial intelligence (hereafter ML/AI) are increasingly ubiquitous in everyday life. Perhaps one of the most impressive implementations of these algorithms has been in the field of open computer vision - whereby a basic webcam can be used to “teach” a computer how to recognize proper exercise form. This seminar will discuss the details of creating ML/AI rehabilitation exercise software, including a discussion of basics of ML/AI including concepts such as collecting training data, making predictions from data and improving predictive accuracy. The seminar will conclude with showing a basic implementation of a full machine-learning workflow that demonstrates how a rehab exercise program can be written, trained on a patient conducting a prescribed exercise program and then used in a home environment to improve patient confidence in conducting exercises properly.
Upon completion participants will be able to define the basic terminology to for machine-learning/AI software development with OpenCV motion detection.
Upon completion participants will be able to describe the steps of implementing a motion tracking ML/AI model.
Upon completion, participants will have created a basic machine learning/AI model of a simple exercise program.
Upon completion, participants will have list of at least three practical implementations of machine-learning/AI in rehabilitation applications.