Longitudinal Evaluation of a 30 Day Therapy Treatment of a Hemiplegic Ankle by Applying a Wearable System and Machine Learning

The amalgamation of wearable systems and machine learning are envisioned to considerably augment the acuity and situational awareness of clinicians for the prescription of a rehabilitation strategy. For example, a therapy regimen for improving the rehabilitation status of a subject’s hemiplegic ankle can be monitored by a wearable system, such as through a smartphone equipped with a software application to function as a wearable and wireless gyroscope platform. Using a machine learning algorithm, such as a support vector machine, various longitudinal phases of the therapy prescription can be differentiated to establish the efficacy of the rehabilitation strategy. Considerable classification accuracy relative to the preliminary first day and final phase of a 30 day longitudinal therapy application has been achieved in conjunction with the application of a smartphone as a wearable system capable of providing a gyroscope signal for subsequent machine learning application through a support vector machine. The implications are for the ability to optimize a rehabilitation strategy from a remote context regarding the clinical team and subject.