2-D Pose Estimation for Remote Thoracic Range of Motion Measurement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current manual goniometers for measuring a patient's range of motion are inaccurate, difficult to perform, prone to variation, and require on-site clinicians, limiting remote health assessments.
Innovation Solution
A 2-D pose estimation method using a camera-based system with AI bot guidance for remotely measuring thoracolumbar rotation range of motion, capturing key-points, and calculating angles without auxiliary sensors or on-site personnel.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual goniometer is used for measuring range of motion, then measurement can be performed with simple device, but measurement accuracy is poor and results show wide variation
Solution Approach 1:
The patent replaces the mechanical goniometer system with an optical-based 2-D pose estimation system using camera(s) and AI algorithms. This substitution eliminates the need for physical contact between clinician and patient while providing more accurate and consistent range of motion measurements through automated key-point tracking and angle calculation.
Solution Approach 2:
The system creates a digital copy of the patient's physical pose through 2-D pose estimation from video or image frames. By identifying key anatomical points and reconstructing the pose digitally, the system enables accurate measurement of range of motion without physical measurement tools, resolving the contradiction between simplicity and accuracy.
2Adaptability or versatility
If manual goniometer is used for measuring range of motion, then on-site clinician presence is required, but this limits remote health assessments
Solution Approach 1:
The system enables patients to perform range of motion assessments independently at home using the camera-based system. The AI bot provides automated guidance through voice or text instructions, and the system automatically captures frames, identifies key-points, and calculates measurements without requiring clinician presence. This self-service capability enables remote assessment while maintaining operational simplicity through automation.
Solution Approach 2:
The patent introduces an AI bot as an intermediary between the clinician and patient. The AI bot delivers instructions to the patient and receives feedback, enabling remote communication and guidance without requiring the clinician to be physically present. This intermediary maintains ease of operation by automating the interaction while enabling remote assessment capability.
3Reliability
If manual goniometer measurement is performed, then clinician can provide real-time guidance, but the process is difficult to perform consistently
Solution Approach 1:
The system implements automated feedback mechanisms where the AI bot provides real-time guidance to patients during the assessment and the system automatically adjusts for proper positioning and movement. This automated feedback ensures consistent measurement protocols are followed, improving reliability while the automation handles the complexity of real-time adjustments that would otherwise require skilled clinician judgment.
Data Source
AI summary
A method of remotely measuring a patient's range of motion using a device having a camera, the device being in data communication with a computer system having a database and an AI bot in communication with the database, the method having: instructing the patient through a stretching procedure; recording a video of the patient as they proceed through the stretching procedure, the video having frames of the patient in a rest pose and a fully stretched pose; capturing the frames of the patient in the rest pose and the fully stretched pose; identifying key-points within the frames; tracing range of movement for the key-points in the frames; calculating a range of motion angle for the stretching procedure by comparing key-point positions between the frames; recording the range of motion angle in the database; and reporting the range of motion angle to the patient or medical personnel.