Multi-layer Image Identification Network for Abnormal Cell Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for identifying abnormal cells in cervical cancer screening are prone to human error due to analyst inexperience, energy-consuming manual adjustments, and visual fatigue, leading to low reading efficiency and accuracy.
Innovation Solution
A method and device that utilize multi-layer images to process and classify single cells and cell clusters, employing a two-stream convolutional neural network to extract both plane and three-dimensional structure information, thereby improving the accuracy of abnormal cell identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual observation and adjustment methods are used by analysts, then the detecting device can be operated with simple equipment, but human errors occur due to inexperience, visual fatigue, and manual energy consumption leading to low reading efficiency and accuracy
Solution Approach 1:
The patent replaces the mechanical manual observation system with an automated image processing system. Multiple layers of cell images are automatically captured, processed through convolutional neural networks, and analyzed to identify abnormal cells. This substitution eliminates manual energy consumption, visual fatigue, and human error while significantly improving both identification accuracy and reading efficiency through automated multi-layer image analysis
Solution Approach 2:
The patent introduces an automated image processing system as an intermediary between the detecting device and the analyst. This intermediary automatically captures multi-layer images, processes them through neural networks, and provides analysis results, thereby eliminating the need for manual observation and adjustment while improving both accuracy and efficiency
2Measurement precision
If multi-layer images are used to extract three-dimensional structure information, then the identification accuracy of abnormal cells is improved, but the complexity of the image processing system increases
Solution Approach 1:
The patent segments the image processing task into distinct components: a first convolutional neural network extracts plane information from individual image layers, while a second convolutional neural network extracts three-dimensional structure information from multiple layers. This segmentation allows the system to handle complex multi-layer image processing through specialized, modular networks, improving identification accuracy while managing system complexity through functional decomposition
Solution Approach 2:
The patent transitions from two-dimensional single-layer image analysis to three-dimensional multi-layer image analysis. By capturing and processing images at multiple focal depths, the system extracts three-dimensional structure information that provides additional diagnostic features for abnormal cell identification, thereby improving accuracy through dimensional expansion
Data Source
AI summary
Methods, apparatus, device, and storage medium for identifying an abnormal cell in a to-be-detected sample are disclosed. The method includes obtaining, by a device, multi-layer images of a to-be-detected sample, the to-be-detected sample comprising a single cell and a cell cluster; obtaining, by the device, multi-layer image blocks of the single cell and multi-layer image blocks of the cell cluster according to the multi-layer images; obtaining, by the device, a first identification result by a first image identification network according to the multi-layer image blocks of the single cell; obtaining, by the device, a second identification result by a second image identification network according to the multi-layer image blocks of the cell cluster; and determining, by the device, whether an abnormal cell exists in the to-be-detected sample according to the first identification result and the second identification result.