Quantitative Movement Testing: Measuring Patient Movements from a Single Smartphone Video

2026-06-01 · Source: Artificial Intelligence · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Computer Vision & Pattern Recognition, Data Science & Analytics · Depth: Advanced, quick

Summary

Quantitative Movement Testing (QMT) is a novel computer vision pipeline designed to extract 3D kinematic biomarkers from standard monocular smartphone video, aiming to balance clinical accessibility with biomechanical accuracy. This system, which utilizes deep learning-based 3D pose-estimation, was validated against gold-standard optical motion capture in 13 healthy controls, achieving strong correlations (r > 0.85) and low mean absolute errors for clinical kinematic metrics. Following calibration, QMT was deployed in two clinical cohorts: a pre- and post-intervention trial for fibromyalgia patients, where it showed high test-retest reliability (r > 0.86), and a 30-day longitudinal monitoring study of chronic sciatica patients, successfully tracking day-to-day movement fluctuations. Despite higher measurement variance in real-world settings, QMT identified group-level differences between healthy controls and sciatica patients using entirely remote recordings, positioning it as a scalable, objective biomarker for disease progression and treatment response.

Key takeaway

For clinical researchers designing remote patient monitoring or intervention trials, consider integrating Quantitative Movement Testing (QMT) to objectively track patient movements. This approach allows you to extract 3D kinematic biomarkers from standard smartphone video, offering a scalable and accessible alternative to traditional lab-based assessments. You can leverage its validated accuracy and reliability to monitor disease progression and treatment response, even in real-world, home-based settings.

Key insights

QMT uses smartphone video and deep learning for accessible, accurate 3D kinematic biomarker extraction in clinical settings.

Principles

• Monocular 3D pose estimation offers scalable assessment.
• Calibration corrects systematic bias for real-world data.
• Remote recordings can identify group-level differences.

Method

The QMT pipeline uses deep learning-based 3D pose-estimation on monocular smartphone video, followed by leave-one-subject-out calibration to correct systematic bias for kinematic biomarker extraction.

In practice

• Track disease progression in chronic pain.
• Monitor treatment response in clinical trials.
• Assess functional ability remotely.

Topics

• Quantitative Movement Testing
• 3D Pose Estimation
• Kinematic Biomarkers
• Remote Patient Monitoring
• Chronic Pain Management
• Computer Vision

Best for: AI Scientist, Computer Vision Engineer, Research Scientist

Related on AIssential

• GloResNet: A lightweight 3D CNN with global topological features for preterm brain injury prediction — Lightweight 3D CNN with global topological features predicts preterm brain injury effectively.
• AI for Interoperability in Health Care: Philips’s Carla Goulart Peron — AI in healthcare augments clinicians to expand access and improve diagnostics, with interoperability identified as the most impactful global capability.
• China has approved the world’s first invasive brain-computer chip—here’s what’s next — China's NEO BCI, less invasive and government-backed, is the first globally approved for commercial use, offering hope for paralysis patients.
• ClinEnv: An Interactive Multi-Stage Long Horizon EHR Environment for Agents — ClinEnv reveals LLMs struggle with sequential, uncertain clinical decisions and information gathering in a multi-stage environment.
• Predicting the risk of colorectal anastomotic leak based on preoperative mapping of the blood supply of the bowel — An AI-driven system predicts colorectal anastomotic leak risk using CT imaging and deep learning for enhanced surgical planning.
• A Minimalist Brain-Computer Musical Interface for Real-Time Emotion-Driven Sonification: System Design and Preliminary Evaluation — Frontal alpha asymmetry proved unreliable for voluntary emotion control in a real-time brain-computer musical interface.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Artificial Intelligence.