Pilot Feasibility Study of a Multi-View Vision Based Scoring Method for Cervical Dystonia

Area of Science:

• Neurology
• Biomedical Engineering
• Medical Imaging

Background:

• Cervical Dystonia (CD) presents with abnormal head and neck movements, necessitating accurate severity scoring for effective treatment.
• Traditional CD assessment methods using protractors or sensors are time-consuming and can impede patient movement.
• The urgency for remote diagnostic tools has increased due to global health concerns.

Purpose of the Study:

• To develop and validate a novel multi-view vision-based method for scoring Cervical Dystonia severity.
• To establish a contactless and low-cost alternative to traditional CD assessment techniques.
• To facilitate remote diagnosis and treatment planning for CD patients.

Main Methods:

• A multi-view vision system was employed to detect keypoint positions from frontal and lateral patient images.
• Head and neck motion angles were calculated from detected keypoints to assess CD severity.
• Scores from the vision-based method were compared against neurologist assessments using the Toronto Western Spasmodic Torticollis Rating Scale (TWSTRS) and analyzed for correlation and absolute accuracy (ICC).

Main Results:

• The vision-based method demonstrated sufficient validity and reliability in scoring CD severity.
• High correlation was observed between the vision-based scores and those determined by a neurologist.
• Intra-class correlation (ICC)(3,1) confirmed the absolute accuracy of the proposed method.

Conclusions:

• The multi-view vision-based scoring method is a valid and reliable tool for assessing Cervical Dystonia severity.
• This contactless, low-cost approach offers a promising solution for remote CD diagnosis and treatment.
• The method addresses the limitations of traditional assessment tools and meets the growing need for telemedicine in managing CD.