Patient Weight Estimation via Depth Sensor Mesh Fitting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Accurate patient weight estimation is challenging in clinical settings, especially for patients with severe conditions, due to the inability to use traditional weighing methods and the time constraints of emergency treatments, leading to moderate accuracy by healthcare workers and potential exposure to higher radiation doses during procedures like chest CT scans.
Innovation Solution
A medical imaging system that uses a depth sensor to capture surface data, fits a patient model as a mesh to the data, extracts shape features, and employs a machine-learned regressor to estimate patient weight, thereby reducing noise and clutter and providing accurate weight estimation for dosing purposes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional weighing methods are used, then accurate weight measurement is achieved, but it cannot be applied to patients with severe conditions and consumes valuable time in emergency treatment
Solution Approach 1:
The patent replaces the mechanical weighing system with an optical-based depth sensing system. A depth sensor captures the patient's body surface geometry, and a fitted patient model extracts weight information from these geometric measurements, eliminating the need for physical contact and mechanical scales that cannot be used with unstable patients.
Solution Approach 2:
The patent introduces a fitted patient model as an intermediary between the raw depth image and weight estimation. The model is fitted to the patient's surface geometry and serves as a mediator that translates complex 3D surface data into meaningful anthropometric measurements, which are then used to estimate weight through regression.
2Extent of automation
If features are extracted directly from depth image, then automation is achieved, but noise and clutter reduce measurement precision
Solution Approach 1:
The fitted patient model acts as an intermediary processing layer between the noisy depth image and the final feature extraction. By first fitting a parametric model to the surface data, the system filters out noise and clutter while preserving essential geometric information, then extracts clean anthropometric features from the fitted model parameters.
3Measurement precision
If manual anthropometric measurements are used, then measurement precision may be improved, but time consumption increases and productivity decreases
Solution Approach 1:
The patent replaces manual measurement processes with an automated computer vision system. Depth sensors capture the patient's geometry, and algorithms automatically fit a patient model and extract anthropometric measurements, eliminating the need for manual tape measurements while maintaining or improving accuracy and significantly reducing time.
Solution Approach 2:
The system performs preliminary actions by capturing the patient's full body geometry with a depth sensor before any measurement extraction occurs. This pre-capture of comprehensive surface data enables multiple measurements to be derived simultaneously from a single scan, rather than taking measurements sequentially as in manual methods.
Data Source
AI summary
For patient weight estimation in a medical imaging system, a patient model, such as a mesh, is fit to a depth image. One or more feature values are extracted from the fit patient model, reducing the noise and clutter in the values. The weight estimation is regressed from the extracted features.