Surgical Image Feedback for Accurate Resection Completion Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing surgical methods lack accurate real-time notification of resection completion during tumor and lymph node removal, leading to potential postoperative recurrence or increased complications.
Innovation Solution
A computer-based system that uses machine learning to analyze surgical images, detecting the state of resection by training with region information and completion labels, and providing real-time notification on a display.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If fluorescence imaging method with ICG-HDL complex is used, then tumor tissue can be distinguished from normal tissue, but the determination of resection completion is left to operator judgment without accurate real-time notification
Solution Approach 1:
The system implements real-time feedback by continuously monitoring surgical images and providing automated notifications about resection completion status. The machine learning model analyzes surgical images in real-time and feeds back completion information to the operator, eliminating reliance on subjective judgment and providing objective, accurate notification of resection status.
Solution Approach 2:
The patent replaces the mechanical/subjective system of operator judgment with an automated machine learning-based image analysis system. The ML model processes surgical images and automatically determines resection completion, substituting human subjective assessment with objective algorithmic analysis that provides consistent and accurate notifications.
2Extent of automation
If machine learning model is trained with surgery images and completion information, then automated notification can be provided, but the accuracy of resection state detection needs improvement
Solution Approach 1:
The system performs preliminary action by training the machine learning model in advance with extensive surgical images and corresponding completion information. This pre-training ensures that when the system is deployed, it already possesses the knowledge and patterns needed to accurately detect resection states, improving detection accuracy before actual surgical use begins.
Solution Approach 2:
The patent uses copying by creating a virtual model of surgical scenarios through training data that replicates actual surgical situations. The machine learning model learns from copied examples of completed and incomplete resections in the training set, enabling it to accurately recognize and notify about resection completion in real surgical scenarios without directly observing them during training.
Data Source
AI summary
Notification of a state of resection of a resection target is accurately provided during surgery. A program according to an embodiment of the present invention causes a computer to: acquire a surgery image in which a surgical site is imaged during surgery for resecting a resection target; detect a state of resection of the resection target from the acquired surgery image by using a machine learning model generated by training with training data in which a surgery image is an input and information related to completion of resection is an output; and notify a state of resection of the resection target, wherein the information related to completion of resection includes region information of an organ that appears in the surgery image when the resection target is resected.