New sensor technology and virtual reality are helping in unsupervised physical therapy

Many chronic disabilities could benefit from physical therapy. Around 75% of the years lived with chronic disability go untreated because there just aren’t enough physical therapists to go around. The number of patients is pacing alongside population growth and aging, and the reporting of cases of severe ailments is increasingly contributing to the issues facing the field.

Why PTs are holistic

There has been growth in the number of sensor-based techniques, such as on-body sensors that track motion are providing autonomy and precision for some sufferers. The minimalist approach to watches and rings largely relies on motion data and therefore lacks the holistic picture of what a physical therapist (PT) pieces together. This includes muscle movement, but also engagement, and tension.

Introducing MuscleRehab

The gap in the muscle-motion treatment plans, or language a PT is trained to understand, has prompted the creation of a physical rehabilitation system, that is unsupervised. Called MuscleRehab, the researchers at MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) along with Massachusetts general hospital (MGH) is designed to help patients seek relief without a PT present.

There are three instances of capture, motion tracking, to capture motion activity, and imaging technique called Electrical Impedance Tomography (EIT) that measures what muscles are doing, and a virtual reality (VR) headset and tracking suit. The VR and suit let the patient watch themselves perform alongside a physical therapist.

Accuracy of exercise increased

With those two conditions, the team was able to compare the accuracy of the exercise and had a professional therapist examine the results. The PT could then explain what muscle groups were supposed to be engaged during each exercise.

With the visualization of both muscle engagement and motion data during an unsupervised exercise routine, instead of just motion alone, the overall accuracy of the exercises improved by 15%.

The tech is popular AI and IoT

These systems are made up of differing types of sensors working in concert in an Internet of Things (IoT) setting. The artificial intelligence algorithms are running inference modeling to determine how each muscle is moving, and from the data determine which muscle groups, such as surface, or deep, are taking on the most beneficial exercise.

This type of imaging of the exercise is similar to being able to use an image, or picture of the exercise and superimpose it on a muscle group. Record the motion and engagement, then have a professional PT look over the results, and recommend exercises.

"We wanted our sensing scenario to not be limited to a clinical setting, to better enable data-driven unsupervised rehabilitation for athletes in injury recovery, patients currently in physical therapy, or those with physical limiting ailments, to ultimately see if we can assist with not only recovery but perhaps prevention, says Junyi Zhu, MIT CSAIL Ph.D. student and lead author on a paper about MuscleRehab. "By actively measuring deep muscle engagement, we can observe if the data is abnormal compared to a patient's baseline, to provide insight into the potential muscle trajectory."

What led to MuscleRehab

Dr. Zhu has been actively digging in to the realm of personal health-sensing devices. He was inspired by the EIT, which measures electrical conductivity of muscles, for a project he did in 2021. That project used non-invasive imaging techniques to create a toolkit for designing and fabrication motion and health sensing devices.

To his knowledge EIT had only been used for monitoring lung function, detecting chest tumors, and diagnosing pulmonary embolisms, but not used in muscle examinations or motion related imaging.

The brains of MuscleRehab

The brains of MuscleRehab is the EIT sensing board. This is accompanied by two straps filled with electrodes which is placed in the suit on the upper thigh, to capture 3D volumetric data. For motion-capturing, the team used OptiTrack, which uses 39 markers and a host of cameras that sense motion at super high frame rates.

This is then shown on the VR screen as highlighted muscle, that actively triggered on the display by becoming highlighted and a darker color then surrounding muscles.

The future of MuscleRehab

The many applications for such a device, are outstanding, and Dr. Zhu and his team are actively working on other muscle group capture techniques.