Title: Wearable Sensors, Smart Phones, and Machine Learning: Impact on Clinical Care and Clinical Trials

Abstract: Machine learning algorithms that use data streams captured from wearable sensors and smart phones have the potential to automatically detect disease symptoms and inform clinicians about the progression or regression of disease. We will discuss on how to design, implement clinical care or research with wearable sensors. The discussion will touch upon on choosing the number and type of sensors to place on an individual, which location on the human body is appropriate to detect symptoms in highly sensitive manner. Furthermore, we will talk about how much data is required or is sufficient to detect symptoms. Does increasing the amount of training data in each individual or adding more individuals lead to improved symptom detection? Which clinical tests or functional behaviors are best suited for symptom detection? We will discuss whether every data analysis requires advanced techniques like convolutional neural networks or other simpler statistical ensembles work to detect symptoms and its progression in each individual. Finally, we will talk about our smart phone technology can be used monitor disease and mobility at home and in the community and inform clinicians remotely the state of the individual under their care.

Bio: Dr. Arun Jayaraman’s work primarily focuses on developing and executing both investigator-initiated and industry-sponsored research in assistive and adaptive technologies to treat physical impairments. He conducts all of his outcomes research using advanced wearable patient monitoring wireless sensors and novel machine learning techniques, in addition to the traditional performance-based and patient-reported outcome measures. He collaborates both nationally and internationally with many academic and industrial organizations and is internationally recognized in the field of wearable technologies.