Automated Burn Assessment Using Image Analysis and Pose Templates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Inexperienced medical providers struggle to accurately assess the severity of burns in patients due to the manual and subjective nature of using standard Lund and Browder charts, leading to significant errors in estimating the total body surface area burned, which can result in suboptimal care.
Innovation Solution
A system and method utilizing a combination of RGB and multispectral images, along with machine learning algorithms, to create a patient-specific enhanced Lund and Browder chart, allowing for accurate determination of burn locations, depth, and percentage of body surface area burned, using pose templates and image analysis to generate both 2D and 3D models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If inexperienced providers manually assess burns using standard Lund and Browder charts, then the assessment process is simple and quick, but the accuracy of TBSA estimation deteriorates significantly with errors of 50%-100%
Solution Approach 1:
The patent replaces the manual mechanical process of visually estimating burn area with automated image processing and computer vision algorithms. The system captures images of the burn patient and uses algorithms to automatically segment and calculate the burned area, eliminating the need for inexperienced providers to manually estimate using charts.
Solution Approach 2:
The patent introduces an intermediary computational system that acts as a bridge between the raw visual information and the final TBSA measurement. This intermediary system processes images through multiple stages including segmentation, landmark detection, and area calculation to provide accurate measurements without requiring expert human judgment.
2Adaptability or versatility
If standard Lund and Browder charts are used with idealized body shapes, then the assessment method is standardized and easy to use, but accuracy deteriorates when patient body shape and size differ from the idealized diagram
Solution Approach 1:
The patent transforms the static, fixed idealized body diagram into a dynamic system that automatically adapts to each patient's actual body shape and size. The image processing system detects the patient's body contours and proportions in the captured images and adjusts the assessment grid accordingly, allowing the same standardized method to accurately accommodate diverse body types.
Solution Approach 2:
The patent applies local adaptation by adjusting the assessment parameters and grid distribution based on the specific body regions and proportions of each patient. Rather than using a uniform idealized chart, the system modifies the local measurement characteristics to match the patient's actual anatomy while maintaining the overall standardized assessment framework.
3Ease of manufacture
If manual coloring and mapping is used on paper or digital LB diagrams, then the process is straightforward and requires minimal equipment, but error introduction increases due to manual mapping between actual patient and idealized shape
Solution Approach 1:
The patent creates a digital copy or replica of the patient's actual body shape and burn patterns from captured images, replacing the manual copying process of coloring on diagrams. The system generates a digital representation that accurately replicates the patient's anatomy and burn distribution, eliminating manual mapping errors while maintaining process simplicity through automated image analysis.
Data Source
AI summary
A system and method of generating an enhanced Lund and Browder chart and total body surface area burn score is described herein. In some embodiments, a plurality of images is obtained from of a patient using a camera system. The images may be taken by aligning the patient's body with pose templates presented on a display of the camera system. The non-skin portions of the images may be removed, and skin analysis performed on the skin portion to determine burn location, coverage, and depth. Further, landmarks may be detected in the images to morph and align the images with the pose templates to obtain standard poses. The plurality of images may be combined and presented in two-dimensional and three-dimensional models with labels and the total surface area burn score.