MMHC

Since the 1980s, motion tracking has been used in animation and biomechanical analysis to accurately follow movement. The suits with ping pong balls and LED lights attached to them remain in use in Hollywood to bring realism to characters in movies like Avatar, as well as in computer gaming.The technology has continued to evolve in the last 10 years. Now, a person’s movements can be tracked without the need for sensors or equipment beyond the camera on a smartphone or laptop. These advances are transforming the delivery of physical therapy home exercise programs.

Sensorlessmotion tracking relies on computer vision and deep learning convolutional neural networks (a type of artificial intelligence) to determine what an object is, what is its position, and monitor its movements over time. Computer vision teaches a computer to process images at the pixel level using special software and algorithms to accurately track objects of interest in an environment. It attempts to be the brain to the phone’s camera, depending on pattern recognition to identify and track moving objects.

Computer vision is already impacting daily life. Tags on photos, augmented reality programs, self-driving cars, tracking consumers in stores, and facial recognition software, all use computer vision.In medicine the technology is being usedto analyze radiographic studies, diagnose skin lesions, and analyze retinal images to name a few. The power of computer vision to speed evaluation and increase accuracy will only grow.

As computing power has increased, the motion tracking abilities of computer vision has greatly advanced, allowing for improved accuracy and precision of measurements, as well as tracking capabilities. Just as self-driving cars have improved with these advancements, so has sensorlessmotion tracking in medicine, negating the need for wearable sensors to the chest and extremities.

We can now use computer vision and sensorlessmotion tracking to monitor a patient’s physical therapy exercises. We track both the quantity of exercises performed and also the quality of the exercises. The data generated (e.g., number of repetitions, days participating in home therapy exercises, knee range of motion, etc.) can all be tracked over time. Tracking data can illustrate improvements but can also discover plateaus or even problems prior to providers or patients themselves recognizing an issue.

The computer vision inputs into the algorithm can trigger cues to the patient such as to bend lower, keep your head up, or keep your knees behind your toes to prevent additional injuries or pain while performing the exercises. The cues can also provide positive feedback like good job or amazing work. The computer vision and algorithm can essentially serve as anautonomous coach to the patient performing the movements. Moreover, sensorless motion tracking still provides the opportunity for clinicians to intervene in real time to advance or modify an exercise program to better personalize care.

Sensorless motion tracking has advanced a great deal over the last 10 years. We expect a similar, if not greater, advancement of the technology over the next 5 years. With such improved accuracy and precision, many more use cases for sensorless motion tracking will be identified, including powerful diagnostic capabilities. At Vori Health, the future is now as we apply sensorless motion trackingtelerehabilitationto support patient engagement, improved clinical outcomes, greater patient convenience, and lower healthcare costs!