A Quantitative Study of Factors Influencing Myasthenia Gravis Telehealth Examination Score

Abstract

IntroductionThe increasing utilization of telemedicine offers an opportunity of video-recording to then perform post-hoc quantitative analysis of all aspects of the patient-examiner interaction. Presently, telemedicine examinations are adapted from the standard clinical visit, but their benefits and deficiencies have not been rigorously assessed.MethodsWe utilized a bank of 54 subjects with myasthenia gravis (MG) each having two telemedicine examinations by neuromuscular experts with recording of the MG Core Examination and the MG Activities of Daily Living. We then applied a broad spectrum of artificial intelligence algorithms from computer vision and speech analysis to natural language processing to generate quantitative metrics of the digital records.ResultsWe successfully developed a technology to assess video examinations. Although the overall MG-CE examination scores were consistent across examiners, individual metrics showed significant variability, with up to a 25 percent variation in scoring within the MG-CE’s range. There was wide range of compliance with MG ADL instructions across examiners.DiscussionWe are able perform digital analysis of neuromuscular examinations and identified variations in outcome measure variations based on the examiner’s instructions, video recording limitations, and the severity of the patient’s disease. Particularly noteworthy is the high standard deviation in scores for examinations of patients with low disease severity.