Glaucoma Diagnosis via Fundus Image Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for diagnosing glaucoma based on the ratio of optic cup to optic disc in color fundus images are inaccurate due to individual differences in body structure, leading to potential misdiagnosis.
Innovation Solution
A method and device for glaucoma auxiliary diagnosis that involves feature extraction and image segmentation of color fundus images to obtain a probability of glaucoma, using a neural network structure to classify pixels and determine the presence of glaucoma by considering multiple dimensions of information, including the entire fundus and optic disc, thereby reducing misdiagnosis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If the ratio of optic cup to optic disc is used to determine glaucoma, then the diagnosis process is simple, but the accuracy is low due to individual differences in body structure
Solution Approach 1:
The patent segments the fundus image into multiple regions including optic disc, optic cup, and other anatomical structures. By dividing the image analysis into distinct segments, the system can extract multiple features beyond just the cup-to-disc ratio, thereby improving diagnostic accuracy while maintaining operational simplicity through automated processing.
Solution Approach 2:
The patent transitions from a single-dimensional approach (cup-to-disc ratio) to a multi-dimensional analysis by extracting multiple features including texture, color, shape, and structural characteristics from different regions of the fundus image. This dimensional expansion enables more accurate glaucoma detection that accounts for individual anatomical variations.
2Device complexity
If only the optic cup to optic disc ratio is considered, then the diagnostic method is simple, but misdiagnosis occurs due to individual structural differences
Solution Approach 1:
The diagnostic system segments the fundus image into multiple anatomical regions and extracts multiple features from each segment. This segmentation approach increases diagnostic reliability by considering multiple anatomical characteristics rather than relying solely on the cup-to-disc ratio, thereby reducing misdiagnosis while the automated process keeps the method relatively simple.
Solution Approach 2:
The patent employs a composite diagnostic approach that combines multiple features (texture, color, shape, structural characteristics) from different image regions to form a comprehensive diagnostic assessment. This composite method improves reliability by integrating multiple lines of evidence, similar to how composite materials combine different properties to achieve superior performance.
3Measurement precision
If multiple features from both optic disc and entire fundus are extracted, then glaucoma recognition accuracy improves, but the processing complexity increases
Solution Approach 1:
The system segments the fundus image into distinct regions (optic disc, optic cup, other areas) and extracts relevant features from each segment. This segmentation strategy manages complexity by organizing feature extraction into modular, region-specific operations while comprehensively capturing multiple diagnostic features that improve glaucoma recognition accuracy.
Solution Approach 2:
The patent implements automated feature extraction and analysis that performs the complex processing tasks without requiring manual intervention. The system self-services by automatically segmenting images, extracting multiple features, and integrating diagnostic information, thereby managing processing complexity through automation while maintaining high diagnostic accuracy.
Data Source
AI summary
A device and method for glaucoma auxiliary diagnosis, and a non-transitory storage medium are provided. The device includes an obtaining unit and a processing unit. The obtaining unit is configured to obtain a color fundus image of a patient. The processing unit is configured to perform feature extraction on the color fundus image to obtain a first feature map. The processing unit is further configured to perform image segmentation on the color fundus image according to the first feature map to obtain an optic disc image in the color fundus image, where the optic disc image corresponds to an optic disc area in the color fundus image. The processing unit is further configured to perform feature extraction on the optic disc image and the color fundus image according to the first feature map to obtain a probability that the patient has glaucoma.