Video-based quantification of hand postural tremor without external references. Integrating postural tremor...

Area of Science:

• Biomedical Engineering
• Neurology
• Medical Imaging

Background:

• Hand postural tremor is a key symptom in neurological disorders like Parkinson's disease.
• Objective quantification of tremor is crucial for diagnosis and treatment monitoring.
• Current methods often require specialized equipment or manual analysis, limiting scalability.

Purpose of the Study:

• To develop and validate an automated method for quantifying hand postural tremor amplitude and frequency from standard videos.
• To assess the method's agreement with manual measurements and its clinical utility in Parkinson's disease patients.

Main Methods:

• Utilized iris-based scaling for accurate size reference within videos.
• Employed deep learning models for 3D pose estimation and precise hand tracking.
• Developed an automated pipeline for tremor quantification.

Main Results:

• The automated method showed excellent agreement with manual measurements of tremor.
• Clinical validation in Parkinson's disease patients demonstrated sensitivity to treatment effects.
• The approach strongly correlated with clinician-provided scores for tremor severity.

Conclusions:

• The proposed method provides an objective and scalable way to assess postural tremor using standard videos.
• This technology, integrated into VisionMD, can enhance the clinical evaluation of movement disorders.
• The non-invasive, calibration-free approach facilitates widespread adoption in clinical settings.