Gait Event Detection and Travel Distance Using Waist-Worn Accelerometers across a Range of Speeds: Automated Approach

Area of Science:

• Biomedical Engineering
• Clinical Biomechanics
• Wearable Technology

Background:

• Accurate measurement of gait clinical features (CFs) is crucial for monitoring mobility in individuals with Duchenne muscular dystrophy (DMD).
• Existing unsupervised methods struggle with the variable gait patterns in DMD patients, hindering accurate mobility assessment.
• Wearable accelerometers offer a promising tool for community-based mobility evaluation, but require robust calibration.

Purpose of the Study:

• To develop and validate a novel calibration method for waist-worn accelerometers to accurately estimate temporospatial gait clinical features (CFs) in children with DMD.
• To assess the method's accuracy across various gait speeds and tasks compared to ground-truth observations.
• To enable reliable, unsupervised computerized measurement of mobility in individuals with progressive ambulatory loss.

Main Methods:

• A novel calibration method combining clinical observation, machine learning for step detection, and regression for stride length prediction was developed.
• Participants included fifteen children with Duchenne muscular dystrophy (DMD) and fifteen typically developing controls (TDs).
• Data was collected using mobile-phone-based waist accelerometers during standardized gait assessments (10 MRW, 25 MRW, 100 MRW, 6 MWT, FW).

Main Results:

• The proposed method demonstrated high accuracy in estimating step counts, distance traveled, and step length for both DMD and TD groups.
• Strong correlations were observed between model predictions and ground-truth data (Pearson's r = -0.9929 to 0.9986, p < 0.0001).
• Mean percentage errors were low: 1.49% for step counts, 1.18% for distance, and 0.37% for step length.

Conclusions:

• A single, calibrated waist-worn accelerometer can accurately measure gait clinical features and estimate travel distance across a range of speeds in children with DMD and their typically developing peers.
• This method provides a reliable tool for objective, community-based mobility assessment in DMD.
• The findings support the use of this calibrated wearable technology for longitudinal monitoring of ambulatory function in DMD.