Striding into Clarity: Wearable Sensor-Driven Estimation of Knee Adduction Moment, Unveiling the Black Box with Sequence-Based Neural Networks and Explainable Artificial Intelligence

Jasmine Liang

Track: Traditional track

Keywords: knee adduction moment, knee osteoarthritis, gait, recurrent neural network, Long Short-Term Memory, wearable sensor, motion capture system, Explainable AI

TL;DR: Our study, leveraging wearable sensors and advanced AI/ML algorithms, accurately predicts knee adduction moment outcomes, with attention weight trends, advanced clinical practice in knee osteoarthritis management

Abstract:

Knee adduction moment during walking has been reported as a sensitive biomechanical marker for predicting the risk of knee osteoarthritis. The traditional method of estimating the knee adduction moment relies on the inverse dynamics approach, primarily limited to laboratory settings due to it relies on specialized equipment and technical expertise, which prevents the clinicians' access to the crucial data. Our study employs wearable sensor technology integrated with advanced Artificial Intelligence and Machine Learning algorithms to predict knee moment outcomes with high accuracy. By analyzing attention weight trends, we establish a significant correlation with knee moment dynamics, validating the reliability of our predictive model. This alignment underscores the biomechanical relevance of our approach, offering promising implications for personalized patient care and clinical practice.

Presentation And Attendance Policy: I have read and agree with the symposium's policy on behalf of myself and my co-authors.

Ethics Board Approval: Yes, we have/will include(d) information about IRB approval or its equivalent, in the manuscript.

Data And Code Availability: No, we will not be making any data and/or code public.

Primary Area: Mechanistic ML approaches for healthcare

Student First Author: Yes, the primary author of the manuscript is a student.

Submission Number: 46