Building a Quantitative Telemedicine Platform for Myasthenia Gravis: Augmenting the Physical Examination

Area of Science:

• Neurology
• Biomedical Engineering
• Artificial Intelligence

Background:

• Myasthenia gravis (MG) poses challenges for remote assessment due to its reliance on physical examination.
• The COVID-19 pandemic accelerated the need for robust telemedicine solutions in healthcare.
• The Myasthenia Gravis Core Exam (MGCE) was developed to standardize telemedicine evaluations for MG.

Purpose of the Study:

• To develop and evaluate a quantitative telemedicine platform for assessing myasthenia gravis.
• To leverage computer vision, signal processing, and AI to augment traditional neurological exams.
• To assess the reproducibility and reliability of the MGCE using machine learning.

Main Methods:

• Development of a telemedicine platform integrating computer vision, signal processing, and AI.
• Application of machine learning algorithms to video and audio data from 52 MG patients.
• Utilizing the Myasthenia Gravis Core Exam (MGCE) with eight digitalizable tasks.
• Performing dual examinations across six centers to assess inter-rater reliability.

Main Results:

• Machine learning extracted clinically relevant features, enabling continuous variation quantitation.
• Identified significant impact of examiner instructions and video quality on reliability.
• Developed a reproducibility of score metric to address inter-rater variability.
• Demonstrated potential for enhanced precision and reduced variability in MG assessments.

Conclusions:

• A digital examination framework can significantly improve MG assessment precision.
• The developed platform supports telemedicine evaluations and reduces physical examination variability.
• This scalable approach can integrate digital biomarkers into neuromuscular disease care and clinical trials.