Medical Image Inference Model Segmentation for Lesion Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In medical image processing, using a single inference model for images with multiple body parts can lead to unsuccessful inferences due to differences in tumor characteristics, and combining multiple models for region-specific inferences can reduce precision at boundary portions.
Innovation Solution
An image processing apparatus that performs first and second inference processing using dedicated and whole-body inference models, respectively, and combines the results to acquire region information on lesion candidates, while also updating region information based on the inference results to improve precision.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single inference model is used for images with multiple body parts, then device complexity is reduced, but measurement precision deteriorates due to differences in tumor characteristics across regions
Solution Approach 1:
The image processing apparatus divides the medical image into multiple regions corresponding to different body parts, and applies different inference models to each region. This segmentation allows each model to specialize in specific anatomical characteristics, thereby improving measurement precision without requiring a single overly complex model to handle all variations.
Solution Approach 2:
The patent implements region-specific inference by assigning different inference models to different body regions based on their unique characteristics. Each region receives processing tailored to its specific tumor characteristics, achieving local optimization of inference quality rather than applying a uniform approach across the entire image.
2Measurement precision
If multiple inference models are combined for region-specific inferences, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The apparatus segments the medical image into multiple regions and applies different inference models to each segment. This segmentation strategy improves precision by matching model characteristics to regional features while managing complexity through modular, region-specific model deployment rather than a single monolithic model.
Solution Approach 2:
The image processing apparatus employs a multi-functional system that can handle different body regions with different inference models within a single integrated framework. This universal approach allows the system to adapt to various regional characteristics while maintaining a cohesive processing architecture.
3Measurement precision
If multiple inference models are used for different regions, then measurement precision is improved, but harmful factors increase due to erroneous detection at boundaries
Solution Approach 1:
The apparatus merges the inference results from multiple region-specific models through result integration processing. This combining approach consolidates the strengths of different models while mitigating their individual weaknesses, particularly at region boundaries where erroneous detections can occur, thereby reducing harmful factors while maintaining improved precision.
Solution Approach 2:
The system incorporates feedback mechanisms where inference results from multiple models are evaluated and integrated, allowing the system to identify and correct erroneous detections at region boundaries. The integration process provides feedback that helps refine the overall inference accuracy by leveraging information from all regional models.
Data Source
AI summary
An image processing apparatus according to an embodiment includes processing circuitry configured to acquire medical image data; acquire, in relation to the medical image data, first region information representing a first region; acquire a first inference result by a first inference processing of applying a first inference model to first medical image data based on the medical image data, and a second inference result by a second inference processing of applying a second inference model to second medical image data based on the medical image data; and acquire, on the basis of the first region information, second region information that is a region satisfying a predetermined condition, from at least regions based on the first inference result or regions based on the second inference result. The processing circuitry is configured to acquire, as the second region information that is the region satisfying the predetermined condition: a region extending from the first region to outside of the first region, the region being among the regions based on the first inference result; or a region extending from the first region to the outside of the first region, the region being among the regions based on the second inference result.