MDA 2025: Digital monitoring tool for myasthenia gravis found feasible
BioDigit MG features sensors that can collect physical activity information
A digital tool developed by BioSensics for remotely monitoring myasthenia gravis (MG) symptoms was found to be feasible in a small clinical study.
BioDigit MG includes various assessments that can be completed on a tablet, along with wearable sensors to collect information about physical activity and posture in daily life.
Ashkan Vaziri, PhD, founder and CEO of BioSensics, discussed the feasibility study’s findings in an presentation titled “Digital Health Technology for Remote Symptoms Monitoring in Myasthenia Gravis” at the Muscular Dystrophy Association (MDA) Clinical & Scientific Conference this month in Dallas. The work was supported by the National Institutes of Health (NIH), the Massachusetts General Hospital, and UCB, which markets the MG therapies Rystiggo (rozanolixizumab-noli) and Zilbrysq.
MG patients experience muscle weakness and fatigue as a result of autoimmune attacks that target proteins important for nerve-muscle communication. Symptoms of the condition include movement problems, speech difficulties (dysarthria), droopy eyelids (ptosis), and breathing issues.
Clinical monitoring is usually done via in-person neurological exams, but these can be time consuming and generally require the expertise of a neuromuscular specialist. They also only capture a single moment in time, which may not accurately reflect how MG affects someone in their everyday life.
“Based on this, we were … motivated to develop a multimodal digital health technology to assess disease symptoms in myasthenia gravis,” Vaziri said.
Monitoring myasthenia gravis remotely
With BioDigit MG, patients complete digital speech assessments, video-based assessments, and other patient-reported outcome measures on a tablet. The tool also incorporates wearable sensors for at-home activity monitoring, along with an at-home breathing assessment.
The clinical study (NCT06277830) involved 20 MG patients to evaluate the technology’s feasibility. The patients wore the sensors for a week before visiting the clinic, where symptoms were assessed with the BioDigit MG tablet, along with standard clinical assessments of disease severity.
Most participants, along with five clinical experts, said they found the tool easy to use, useful, and well suited for their needs.
“Both experts and patients … rated the technology highly and were motivated to use them,” Vaziri said.
Data from 219 speech tasks and 119 video tasks were collected and analyzed using the company’s BioDigit Speech and BioDigit Video tools. Analyzed speech data included aspects such as speech intelligibility, missing words, pausing, and pitch.
The scientist said it was difficult to evaluate how well the digital assessments correlated with clinician-rated measures of speech because only two people had dysarthria, but indicated available data did suggest correlations.
Vaziri said BioDigit Speech can use artificial-intelligence driven models to help diagnose neurological disorders based on specific speech patterns collected with the digital tool. He pointed to analyses conducted in people with amyotrophic lateral sclerosis and other neurological diseases, where the algorithm had a high accuracy for predicting each disease.
The BioDigit Video tool was used to assess features such as facial characteristics, hand and finger movements, and gait. Results showed that digitally-assessed parameters related to eye aspect ratio, which evaluates how wide open the eye is, significantly correlated with clinician-rated measures of ptosis.
The movement sensors collected a range of data, including aspects related to posture, such as sitting, standing, and walking; going from sitting to standing; and walking, such as step counts and rhythm. A number of these parameters correlated with standard clinical assessments of MG symptoms and fatigue.
The researchers are now working to further validate the technology in larger trials.
“As the next step, we have started … running a large, multi-site longitudinal study using our system,” Vaziri said.
Called the BioDigit MG study, it will involve 100 patients who will be monitored for a year, including clinic visits every three months, digital assessments every two weeks, and continuous sensor data collection. Along with establishing BioDigit MG as a remote monitoring tool, the study also seeks to enable “early prediction of disease worsening,” according to Vaziri.
Another longitudinal natural history study (NCT06630650) being conducted at NIH is employing the tool to monitor people with congenital myasthenic syndromes, a group of genetic disorders that clinically overlap with MG, but have different underlying causes. BioSensics is also investigating its digital monitoring tools across a range of other neurodegenerative and neuromuscular diseases.
About the Author
