Development and validation of an automated Trunk Impairment Scale 2.0 scoring system using rule-based classification

Area of Science:

• Rehabilitation Medicine
• Biomedical Engineering
• Movement Science

Background:

• Stroke survivors often experience trunk motor impairment, affecting balance and coordination.
• Manual assessment of trunk motor function using the Trunk Impairment Scale Version 2.0 (TIS 2.0) by physiotherapists can be time-consuming and subjective.
• There is a need for objective, reproducible, and efficient methods to assess trunk motor impairment post-stroke.

Purpose of the Study:

• To develop and validate an automated scoring system for the TIS 2.0.
• To provide a standardized and reproducible alternative to manual TIS 2.0 assessments performed by physiotherapists.
• To evaluate the performance of the automated system in measuring trunk motor impairment in stroke survivors.

Main Methods:

• Collected motion data from actors simulating stroke using the OpenCap video motion capture system for metric development.
• Established cut-off values for a rule-based classification system.
• Tested the automated system on 19 stroke survivors, comparing its scores to manual assessments by nine physiotherapists.

Main Results:

• The automated system demonstrated excellent to outstanding discriminant abilities for the dynamic sitting balance subscale (AUC ≥ 0.717).
• The coordination subscale showed discriminant abilities ranging from poor to outstanding (AUC ≥ 0.667).
• The automated TIS 2.0 scores achieved an average agreement of 71.1% with physiotherapists' scores and showed moderate correlation (R=0.526, p<0.05) with sum of mode-determined task scores.

Conclusions:

• The proposed automated scoring system for the TIS 2.0 is a valid and reliable tool for assessing trunk motor impairment in stroke survivors.
• The automated system offers a reproducible and standardized alternative to manual assessments, potentially enhancing rehabilitation efficiency.
• This technology has the potential to improve the consistency and accessibility of TIS 2.0 evaluations.