Convolutional Neural Network for Burn Wound Depth Assessment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for assessing burn wound depth, such as clinical assessment and laser Doppler imaging, are unreliable, leading to inaccurate determination of burn wound severity, resulting in unnecessary surgery, prolonged treatment, and increased scarring or disfigurement.
Innovation Solution
A system employing a convolutional neural network classifier that evaluates nodes and edges in ultrasound imaging videos of burn wounds, comparing them to historical data to accurately distinguish between second and third degree burns, using a database of known burn wounds and providing a depth assessment based on similarity and probability analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If clinical assessment is used to determine burn wound depth, then the method is simple and widely available, but the accuracy is low (50-75% accuracy)
Solution Approach 1:
The patent introduces an artificial intelligence system as an intermediary between the clinical assessment and the final diagnosis. The AI system processes ultrasound images and provides objective measurements of burn depth, acting as a mediator that enhances the accuracy of the simple clinical assessment method without requiring complex surgical intervention.
Solution Approach 2:
The patent replaces the subjective mechanical visual assessment with an automated image processing system using artificial intelligence. The AI system analyzes ultrasound images to objectively determine burn depth, substituting the unreliable human visual inspection with a more accurate automated analysis method.
2Measurement precision
If laser Doppler imaging is used to assess burn wound depth, then the measurement precision is improved, but the device complexity and difficulty of interpretation increase
Solution Approach 1:
The patent uses ultrasound imaging to create a copy or representation of the tissue structure that can be analyzed without requiring complex light-based detection systems. The AI system processes these ultrasound copies to determine burn depth, avoiding the need for complex laser Doppler equipment while achieving similar diagnostic accuracy.
Solution Approach 2:
The patent changes the physical parameter used for assessment from light-based blood flow detection (laser Doppler) to sound-based tissue structure imaging (ultrasound). This parameter change simplifies the equipment requirements while providing complementary information about burn depth through AI analysis of tissue echogenicity and structural changes.
3Measurement precision
If laser Doppler imaging is used, then measurement precision is improved, but the reliability decreases due to interference from pigments, edema, and vasoactive medicines
Solution Approach 1:
The patent segments the assessment into multiple independent parameters using ultrasound imaging, including tissue echogenicity, structural integrity, and layer differentiation. By analyzing multiple segmented features rather than relying on a single blood flow parameter, the system achieves more reliable results that are not affected by vasoactive medicines, edema, or pigments that interfere with light-based detection.
Solution Approach 2:
The AI system acts as an intermediary that processes ultrasound images to extract multiple independent features for burn depth assessment. This intermediary approach allows the system to bypass the limitations of direct blood flow measurement and use alternative tissue characteristics that are not affected by patient conditions such as edema, pigments, or vasoactive medication.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach improves the accuracy of burn wound depth assessment, reducing unnecessary surgery for second degree burns and ensuring appropriate treatment for third degree wounds, thereby enhancing treatment outcomes.
Implementation Method 1
receiving an imaging video of the burn wound collected by an ultrasound transducer
Data Source
AI summary
A method for evaluating the depth of a burn wound in a subject comprises (a) receiving an imaging video of the burn wound collected by an ultrasound transducer, (b) evaluating nodes and edges connecting the nodes in a plurality of layers of the burn wound in a plurality of frames of the imaging video, with a classifier system, the classifier system comprising a convolutional neural network that compares the nodes and edges of the burn wound with nodes and edges in a plurality of layers of a plurality of historical burn wounds of known depth, and (c) assigning a depth to the burn wound, the assigned depth being an average of depths associated with a plurality of the historical burn wounds having a predetermined minimum similarity of nodes and edges to the nodes and edges of the burn wound.