Quantification of Tongue Motor Dysfunction in Amyotrophic Lateral Sclerosis Using a Smartphone-Based Task and Deep...

Area of Science:

• Neurology
• Biomedical Engineering
• Artificial Intelligence in Medicine

Background:

• Bulbar dysfunction is a significant complication in amyotrophic lateral sclerosis (ALS).
• Objective assessment of tongue movements is crucial for monitoring bulbar function in ALS patients.
• Current methods for assessing bulbar function may not capture subtle impairments.

Purpose of the Study:

• To develop and validate a simple, smartphone-based task for objective tongue movement assessment in ALS.
• To examine the association between quantified tongue movement metrics and clinical variables in ALS.
• To explore the utility of this technology for remote monitoring and disease management.

Main Methods:

• 37 ALS patients and 20 controls performed a smartphone-recorded tongue lateralization task.
• A deep-learning U-Net++ model was employed for image segmentation and feature extraction.
• Tongue movement frequency and amplitude were quantified and correlated with clinical measures (ALSFRS-r, fasciculations, jaw jerk, spasticity).

Main Results:

• The U-Net++ model demonstrated robust segmentation performance.
• ALS patients exhibited significantly lower tongue movement frequency compared to controls (0.14 vs. 0.40, p < 0.001).
• Normalized tongue movement frequency correlated with dysarthria (p = 0.003) and speech performance (p = 0.009), along with tongue spasticity (p = 0.015).

Conclusions:

• The smartphone-based method offers an objective, minimally invasive measure of bulbar function in ALS.
• This approach correlates with clinical assessments and can detect subtle impairments.
• This technology shows promise for remote patient monitoring and improved ALS management.