OpenCap: Human movement dynamics from smartphone videos

Area of Science:

• Biomechanics
• Human Movement Analysis
• Wearable Technology

Background:

• Quantifying human movement dynamics is crucial for predicting injury risk and disease progression.
• Current methods are limited in large-scale studies and clinical practice due to high costs, time, and expertise requirements.

Purpose of the Study:

• To introduce and validate OpenCap, an open-source platform for computing human movement kinematics and dynamics.
• To demonstrate OpenCap's ability to democratize access to advanced biomechanical analysis.

Main Methods:

• Utilizes smartphone videos with pose estimation algorithms to identify body landmarks.
• Employs deep learning and biomechanical models for 3D kinematics estimation.
• Applies physics-based simulations to determine muscle activations and musculoskeletal dynamics.
• Features a cloud-based web application for automatic data processing and visualization.

Main Results:

• OpenCap accurately predicts key dynamic measures, including muscle activations, joint loads, and joint moments.
• Demonstrated a 25x speed increase and <1% cost reduction compared to traditional lab-based methods in a 100-subject study.
• Validated the platform's ability to screen for disease risk and evaluate intervention efficacy.

Conclusions:

• OpenCap significantly lowers the barrier to entry for human movement dynamics analysis.
• The platform facilitates the integration of biomechanical metrics into large-scale research, clinical trials, and practice.
• OpenCap empowers faster, more cost-effective assessment of musculoskeletal health and movement patterns.