EE Seminar: DEVELOPMENT OF A NEW DYNAMIC ALGORITHM FOR EXERCISE MONITORING AND PROGRAMING
Speaker: Nir Minerbi
M.Sc. student under the supervision of Prof. Mickey Scheinowitz
Monday, February 25th, 2019 at 15:00
Room 011, Kitot Bldg., Faculty of Engineering
DEVELOPMENT OF A NEW DYNAMIC ALGORITHM FOR EXERCISE MONITORING AND PROGRAMING
Abstract
Background- Many studies have found that sedentary lifestyle is associated with an increased risk of diabetes, obesity, heart disease, cancer and premature death. These studies show that maintaining high physical fitness level does not prevent the damage caused by prolonged sitting [1]. Despite having a large variety of sport technologies, very few provide real-time feedback and neither were clinically proven on cardiac patients.
Goal- To create a dynamic algorithm for monitoring and recommending effective exercise recommendations, both as part of the training program and as part of leisure time physical activity.
Methods- Unlike most of applications on market today, whom handle daily activity and training program separately, this algorithm aims to analyze both activity types and to make bi-directional adjustments for serving the main goal: user’s fitness improvement and health outcome. For maximizing the efficiency of the application, the algorithm will adapt to the capabilities and needs of the user and will feed him back according to his progress. The purpose of the current thesis was therefore to develop a virtual personal trainer, which helps the user to enhance physical fitness, and reduce as much as possible the damage associated with physical inactivity during daily routine. In order to test the efficiency of the newly developed algorithm, it was tested under a clinical experimental setting in the cardiac rehabilitation institute, Sheba Medical Centre.
Results- the results show that the algorithm achieved much better results comparing the control group. In addition, statistical comparison to meta-analysis of 31 similar researches showed that the developed algorithm showed up to 65% higher fitness improvement results among the patients participated in the experiment than other rehabilitation processes. Conclusions- The newly developed algorithm was proven to be as good as human guidance and can potentially replace human accompaniment in the cardiac rehabilitation setting, in the future.
