FNA Biopsy Image Analysis for Real-Time Tissue Sufficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The FNA method for tissue extraction during endoscopic procedures requires a subsequent endoscopy if tissue sufficiency is uncertain, and real-time determination of tissue adequacy is challenging due to the absence of continuous clinical pathologist presence.
Innovation Solution
A method and device utilizing deep learning technology to analyze tissue samples via convolutional neural networks for real-time assessment of tissue sufficiency and cancer-probable cell identification using image data from suspected cancer lesions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If FNA method is used for tissue extraction, then tissue can be acquired during endoscopic procedure, but real-time determination of tissue sufficiency is difficult requiring repeat endoscopy
Solution Approach 1:
The patent applies preliminary action by performing rapid on-site evaluation of the tissue sample immediately after FNA extraction. The evaluation system analyzes cellular morphology, nuclear features, and tissue architecture in real-time to determine tissue adequacy before the procedure concludes, preventing the need for repeat endoscopy and eliminating the time loss associated with delayed assessment.
Solution Approach 2:
The patent introduces an intermediary evaluation system that acts as a mediator between tissue extraction and final diagnosis. This system includes specialized staining protocols, microscopic evaluation equipment, and trained cytopathologists who assess tissue adequacy on-site, serving as an intermediate step that provides real-time feedback without requiring repeat procedures.
2Measurement precision
If clinical pathologist performs ROSE process to determine tissue sufficiency, then tissue adequacy can be assessed, but clinical pathologist cannot always be present in laboratory
Solution Approach 1:
The patent applies segmentation by dividing the tissue evaluation process into distinct modular components: initial rapid screening using automated image analysis, intermediate evaluation by cytotechnicians, and final confirmation by pathologists. This segmented approach allows multiple levels of assessment to occur sequentially, maintaining high accuracy while improving operational availability through tiered expertise deployment.
Solution Approach 2:
The patent employs copying by creating digital replicas of tissue samples through high-resolution imaging and creating virtual slides that can be reviewed remotely. These digital copies allow pathologists to assess tissue adequacy from any location, eliminating the constraint of physical presence in the laboratory while maintaining diagnostic accuracy through detailed virtual examination.
3Productivity
If deep learning technology is used to analyze tissue samples, then real-time cell examination is enabled, but device complexity increases
Solution Approach 1:
The patent applies preliminary action by implementing automated preprocessing steps that prepare tissue images for deep learning analysis before the main examination. This includes automated staining optimization, image normalization, artifact removal, and feature enhancement that occur automatically before the deep learning model processes the data, enabling real-time analysis without manual intervention.
Solution Approach 2:
The patent introduces intermediary components between the tissue sample and the deep learning system, including automated image capture devices, digital slide scanners, and preprocessing software that convert physical tissue samples into standardized digital formats suitable for AI analysis. These intermediaries bridge the gap between biological samples and computational algorithms, managing system complexity through specialized interface layers.
Data Source
AI summary
Disclosed is a method for examining cells including: acquiring image data of cells acquired from a region suspected to be a cancer lesion; analyzing the image data; determining whether an amount of the cells is sufficient based on the analyzed result; and determining whether cancer-probable cells are included in the cells based on the analyzed result and the determined result; wherein the cells are acquired through Fine Needle Aspiration (FNA).