JMIR Publications

ABSTRACT

Background:

Telemedicine practice for neurological diseases has grown significantly during the COVID-19 pandemic. It is unclear however if diagnosis and monitoring of neuromuscular disorders are done with the same level of quality and sensitivity than neurologist practice during face-to-face examination. Telemedicine offers a natural approach to the digitalization of the examination and enhances access to modern computer vision and artificial intelligence processing to annotate and quantify the examination in a consistent and reproducible manner. The Myasthenia Gravis Core Exam was recommended for telemedicine evaluation of patients with myasthenia gravis and involving different physical exercises.

Objective:

Our goal is to take accurate and robust measurements during these exercises that remove the human bias, and can be used to support the clinical study. Our two objectives are to improve: (i) the workflow efficiency by making the data acquisition and analytics fully automatic and (ii) the examination quality by removing human observation bias.

Methods:

The core examination tests require two broad categories of processing: first the computer vision algorithms to analyze video focusing on eye or body motions. Second, the analysis of the voice signal, that requires a completely different category of signal processing methods. In this way, we are providing an algorithm toolbox to assist clinician with the myasthenia gravis core examination.

Results:

The digitalization and control of quality of the core exam is advantageous and let the medical examiner concentrate on the patient instead of managing the logistic of the test. This approach showed the possibility of a standardized data acquisition during telehealth sessions and provided real-time feedback on the quality of the metrics the medical doctor is looking for. Overall, our new telehealth platform provides a submillimeter accuracy on ptosis and eye motion . On top of this, the method shows good result in monitoring muscle weakness proof on a continuous analysis rather that pre and post exercise subjective observations.

Conclusions:

Based on the new data set that our method provides, one should investigate further if the core exam classification should be revisited to take into account some of the new metrics that our algorithm can provide. As a matter of fact, the time dynamic component of muscle weakness seems an important factor in quality of life. We provide a proof of concept with the Myasthenia Gravis Core Examination but, the method and tools developed may apply to many neurological disorders and have great potential to improve clinical care.