Head tremor in cervical dystonia: Quantifying severity with computer vision.

Area of Science:

• Neurology
• Biomedical Engineering
• Medical Technology

Background:

• Head tremor (HT) is a prevalent symptom in cervical dystonia (CD), often assessed subjectively.
• Objective and automated methods for quantifying HT severity are needed.

Purpose of the Study:

• Develop Computational Motor Objective Rater (CMOR), a computer vision system, to objectively quantify head tremor.
• Assess oscillatory and directional aspects of HT from video recordings.
• Validate CMOR's convergent validity with established clinical rating scales.

Main Methods:

• Analyzed video recordings of 93 participants with isolated CD and HT.
• Used CMOR to evaluate peak power, frequency, and directional dominance of head tremor.
• Employed Spearman's correlation to compare CMOR measures with clinical ratings.

Main Results:

• CMOR quantified head tremor power and peak frequency (1.95 Hz).
• Pitch was the dominant tremor axis for 50% of participants, with significant contributions from secondary and tertiary axes.
• CMOR's HT severity measure showed a significant positive correlation with the Toronto Western Spasmodic Torticollis Rating Scale-2 (rho=0.54, p<0.001).

Conclusions:

• A novel, objective method for measuring head tremor severity in cervical dystonia has been developed.
• CMOR utilizes conventional video recordings to quantify complex HT characteristics.
• The system demonstrates convergent validity with clinical assessments, supporting its utility in research and practice.