Artificial Neural Network Learns Clinical Assessment of Spasticity in Modified Ashworth Scale

Abstract

Objective:To propose an artificial intelligence (AI)-based decision-making rule in modified Ashworth scale (MAS) that draws maximumagreement from multiple human raters and to analyze how various biomechanical parameters affect scores in MAS. Design:Prospective observational study. Setting:Two university hospitals. Participants:Hemiplegic adults with elbow flexor spasticity due to acquired brain injury (NZ34). Intervention:Not applicable. Main Outcome Measures:Twenty-eight rehabilitation doctors and occupational therapists examined MAS of elbow flexors in 34 subjects withhemiplegia due to acquired brain injury while the MAS score and biomechanical data (ie, joint motion and resistance) were collected. Ninebiomechanical parameters that quantify spastic response described by the joint motion and resistance were calculated. An AI algorithm (orartificial neural network) was trained to predict the MAS score from the parameters. Afterwards, the contribution of each parameter fordetermining MAS scores was analyzed. Results:The trained AI agreed with the human raters for the majority (82.2%, Cohen’s kappa=0.743) of data. The MAS scores chosen by the AIand human raters showed a strong correlation (correlation coefficient=0.825). Each biomechanical parameter contributed differently to thedifferent MAS scores. Overall, angle of catch, maximum stretching speed, and maximum resistance were the most relevant parameters thataffected the AI decision. Conclusions:AI can successfully learn clinical assessment of spasticity with good agreement with multiple human raters. In addition, we couldanalyze which factors of spastic response are considered important by the human raters in assessing spasticity by observing how AI learns theexpert decision. It should be noted that few data were collected for MAS3; the results and analysis related to MAS3 therefore have limitedsupporting evidence.