Medical Image Inspection Part Identification Using Subject Pose Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing medical imaging systems struggle with limited accuracy in identifying inspection parts due to reliance on template matching, which is constrained by specific environmental conditions and positional data.
Innovation Solution
An information processing apparatus and method that utilize a camera and ultrasound probe to capture external appearance images, combined with machine learning to predict the position and orientation of a subject, enabling accurate identification of inspection parts through image analysis and automated adjustment of the camera's position and orientation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If template matching is used to identify inspection parts, then the system can operate with simple equipment, but the identification accuracy is limited by environmental conditions and positional data
Solution Approach 1:
The patent replaces the mechanical template matching system with a machine learning-based image recognition system. The machine learning model processes external appearance images to predict subject position and orientation, and identifies inspection parts without relying on pre-defined templates or specific environmental conditions, thereby improving both accuracy and adaptability
Solution Approach 2:
The system changes the approach from fixed template parameters to dynamic machine learning predictions. By using machine learning to predict subject position and orientation from external appearance images, the system adapts to varying environmental conditions and achieves higher identification accuracy across different scenarios
2Productivity
If manual identification of inspection parts is performed, then the system requires less complex processing, but the diagnostic efficiency is reduced
Solution Approach 1:
The system implements automated identification of inspection parts using machine learning. The machine learning apparatus autonomously processes external appearance images, predicts subject position and orientation, and identifies inspection parts without requiring manual intervention, thereby significantly improving diagnostic efficiency while the automated nature handles the complexity internally
Solution Approach 2:
The system performs preliminary prediction of subject position and orientation from external appearance images before conducting the actual inspection part identification. This preliminary action prepares the system for accurate and efficient identification, enabling automated operation that improves diagnostic efficiency
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An information processing apparatus comprising: acquire means for acquiring a first image including at least a portion of an inspection device, and a second image including at least a portion of a subject; first prediction means for predicting a position of the inspection device based on the first image; and second prediction means for predicting position/orientation information regarding the subject based on the second image. The apparatus comprises part identification means for identifying an inspection part of the subject based on prediction results of the first and second prediction means. Based on a learning model trained in advance using a plurality of images of training data including subjects similar to the subject, the second prediction means is arranged to output the position/orientation information regarding the subject.