An Upper Limb Rehabilitation Exercise Status Identification System Based on Machine Learning and IoT

Home-Based Rehabilitation of the Shoulder Using Auxiliary Systems and Artificial Intelligence: An Overview

“…The adoption of rehabilitation systems integrating machine-learning techniques has further accelerated in the wake of the global COVID-19 outbreak. These systems offer several advantages, including remote access to patients’ exercise information without the need for direct physical contact, enabling health professionals to monitor their patients’ progress at any time [ 75 ]. This remote supervision capability not only ensures continuous care but also allows health professionals to simultaneously oversee the progress of multiple patients, significantly enhancing their productivity.…”

“…These include the superficial electromyography (EMG) of muscles in the shoulder and arm [ 39 ], range of movement, movement speed [ 102 ], angular acceleration, and angular velocity [ 103 ]. Additionally, factors such as exercise duration and completion status [ 75 ] can also be considered. By obtaining these parameters, evaluations can be based on performance-based impairment indices, such as the Fugl-Meyer motor scale [ 104 ], the Wolf motor arm test [ 105 ], the shoulder pain and disability index [ 106 ], or the modified motor assessment scale [ 107 ].…”