Medical Image Processing Apparatus for Automated Learning Data Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing medical image processing systems face challenges in efficiently extracting appropriate learning data for generating accurate algorithms, as they require extensive labor to identify medical image data that meets specific criteria such as image quality and region details, especially for algorithms like brain segmentation from T1-weighted images.
Innovation Solution
A medical image processing apparatus that performs image analysis on collected data to extract medical image data with attributes common to teaching data used for generating a learned model, reducing the labor required for data extraction by utilizing a VNA, AI service server, and processing circuitry to automate the selection and processing of relevant data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual extraction of learning data is performed from medical image data, then data selection accuracy can be ensured, but labor time and operational complexity increase significantly
Solution Approach 1:
The system enables self-service by allowing the medical image processing apparatus to automatically extract learning data candidates based on predetermined conditions stored in its own storage unit, without requiring external manual selection. The apparatus compares image data against stored conditions and autonomously identifies suitable learning data, reducing operational complexity while maintaining accuracy through systematic automated evaluation
Solution Approach 2:
The system applies preliminary action by pre-storing extraction conditions, criteria, and parameters in the storage unit before actual learning data extraction is needed. These predetermined conditions include image quality thresholds, anatomical region specifications, and other selection criteria that guide the automated extraction process, enabling efficient and accurate data selection without manual intervention during operation
2Reliability
If extensive manual review of medical image data is conducted to meet algorithm criteria, then learning data quality improves, but processing time increases
Solution Approach 1:
The system replaces the mechanical manual review process with an automated image processing and comparison system. The processing circuitry automatically retrieves image data, compares it against predetermined conditions using computational algorithms, and identifies learning data candidates based on objective criteria such as image quality metrics and anatomical region detection, significantly reducing processing time while maintaining consistent quality standards
Solution Approach 2:
The system utilizes parameter changes by evaluating multiple image parameters simultaneously (such as image quality scores, anatomical region presence, contrast levels, and resolution metrics) against predetermined thresholds. This multi-parameter automated evaluation efficiently identifies high-quality learning data without the time-consuming sequential manual review process, maintaining reliability through comprehensive parameter assessment
3Productivity
If automated extraction methods are used, then processing efficiency increases, but extraction accuracy may deteriorate
Solution Approach 1:
The system implements feedback by storing predetermined extraction conditions and criteria in the storage unit that are continuously referenced during the automated extraction process. The processing circuitry compares extracted data against these stored conditions and can adjust or refine extraction parameters based on the results, ensuring that automated extraction maintains high accuracy by continuously validating against established standards rather than using fixed rigid parameters
Data Source
AI summary
A medical image processing apparatus according to an embodiment includes a processing circuit. The processing circuit specifies teaching data used for generation of a learned model. The processing circuit performs image analysis with respect to pieces of collected medical image data. The processing circuit extracts medical image data having an attribute common to the teaching data of the learned model from the pieces of medical image data based on an analysis result of the image analysis, as a candidate of the teaching data of the learned model.