Machine Learning Algorithm for Surgical Microscope Tissue Marking Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In surgical microscopy, there is a need to improve the accuracy of marked tissue sections, as false positives and false negatives can lead to incorrect resection of tissue during surgeries.
Innovation Solution
A system and method for training a machine-learning algorithm using microscope images from a surgical microscope, which adjusts the algorithm to correct falsely marked sections of tissue, by utilizing annotations and corrected images as training data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If fluorophores or markers are supplied to tissue sections to enable visualization during surgery, then relevant sections appear coloured and can be identified, but false markings occur leading to incorrect identification of tissue sections
Solution Approach 1:
A machine learning algorithm is introduced as an intermediary between the fluorescence microscope image and the surgeon's decision-making process. The algorithm analyzes the marked sections and provides correction information by comparing the fluorescence signal with corresponding visible light images, thereby mediating to reduce false positives and false negatives in tissue marking
Solution Approach 2:
The system implements feedback by providing corrected marking information back to the surgeon through the display unit. The machine learning algorithm continuously analyzes new fluorescence images and compares them with reference data, providing real-time feedback to correct previous marking errors and improve subsequent marking accuracy
2Measurement precision
If machine learning algorithms are trained using training data including annotations and corrected images, then the algorithm can correct marked sections of tissue, but the system complexity increases
Solution Approach 1:
The machine learning algorithm is trained in advance using pre-prepared training data that includes annotated fluorescence images and corresponding corrected images. This preliminary training action enables the algorithm to learn the patterns of false markings and correction strategies before actual surgical use, reducing the need for complex real-time processing during surgery
Solution Approach 2:
The system uses copies of training data in multiple formats (annotated images, corrected images, and corresponding visible light images) to train the machine learning algorithm. These copied and transformed versions of the original data enable comprehensive training without requiring additional physical resources during the actual surgical procedure
Data Source
AI summary
A system for training a machine-learning algorithm includes one or more processors, and one or more storage devices. The system is configured to receive training data. The training data includes microscope images from a surgical microscope obtained during a surgery. The microscope images show tissue. The system is further configured to adjust the machine-learning algorithm based on the training data to obtain a trained machine-learning algorithm, such that the trained machine-learning algorithm corrects marked sections of tissue in a microscope image of the microscope images, and provide the trained machine-learning algorithm.