Deep Learning PCCT Image Viewer for Guided Diagnostic Review

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional image viewers fail to leverage the high-resolution and improved image quality of PCCT images, hindering effective utilization of their advantages in clinical workflows.

Innovation Solution

A deep learning-based PCCT image viewer that utilizes a language model for clinical task selection and machine learning models for medical imaging analysis, providing a guided review of PCCT images through whole-image, anatomical, and pathological compartmental reviews, optimizing the clinical workflow by focusing on relevant findings and enhancing diagnosis efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If conventional image viewers are used to view PCCT images, then the imaging process is simple and compatible with existing workflows, but the high-resolution advantages and improved image quality of PCCT images cannot be effectively utilized

Engineering Contradiction:
Improveability to utilize PCCT image advantagesVSAvoidviewer system complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent segments the image review process into multiple levels: whole-image review, compartmental review (anatomical regions), and findings review (pathological regions). This segmentation allows the system to leverage PCCT's high-resolution capabilities selectively in different review stages without requiring the entire system to be uniformly complex, thus improving adaptability while managing device complexity.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system performs preliminary actions by automatically generating compartmental and findings reviews before the radiologist conducts the full diagnostic process. Machine learning models pre-identify anatomical regions and pathological findings, allowing the radiologist to focus on confirming and interpreting these pre-processed results, thereby utilizing PCCT's high resolution efficiently without manual effort.

Inventive Principle:
Principle #10Preliminary action

2Reliability

If manual review of all PCCT images is performed, then comprehensive diagnosis is achieved, but time consumption increases significantly

Engineering Contradiction:
Improvediagnostic accuracyVSAvoidreview time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The system performs preliminary identification of anatomical compartments and pathological findings using machine learning models before the radiologist conducts comprehensive review. This preliminary action filters and organizes the high-resolution PCCT data, allowing radiologists to focus their expertise on confirming and interpreting key findings rather than manually scanning all images, thus maintaining diagnostic accuracy while reducing time consumption.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent introduces machine learning models as intermediaries between the PCCT imaging system and the radiologist. These models automatically perform initial analysis, generating compartmental and findings reviews that serve as intermediates, allowing the radiologist to efficiently verify and interpret results without directly examining every raw image, thereby balancing reliability and time efficiency.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Loss of information

If high-resolution PCCT images are displayed in full detail, then diagnostic information is maximized, but the complexity of image navigation and analysis increases

Engineering Contradiction:
Improvediagnostic information retentionVSAvoidimage navigation ease
Core Design Contradiction:
Loss of informationVSEase of operation

Solution Approach 1:

The patent segments the high-resolution PCCT images into manageable compartments based on anatomical regions and further into findings based on pathological areas. This segmentation preserves all diagnostic information in the original high-resolution images while organizing them into navigable sections, allowing radiologists to easily navigate to specific regions of interest without being overwhelmed by the complexity of viewing entire high-resolution datasets.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system adds a new dimension of organization by creating a hierarchical structure: whole-image level, compartmental level (anatomical regions), and findings level (pathological areas). This dimensional organization allows radiologists to navigate through images at different levels of detail, preserving all diagnostic information while significantly improving ease of operation through structured access to high-resolution data.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

Data Source

PatentEP4607524A1Deep learning based PCCT image viewer
Publication Date: 2025.08.27 SIEMENS HEALTHINEERS AG
  • EP4607524A1 patent drawingFigure 1
  • EP4607524A1 patent drawingFigure 2
  • EP4607524A1 patent drawingFigure 3

AI summary

Systems and methods for generating a guided review of the one or more input medical images are provided. One or more input medical images of a patient and text-based patient data of the patient are received. One or more clinical tasks are identified based on the text-based patient data using a language model. One or more machine learning based models are selected based on the one or more identified clinical tasks. One or more medical imaging analysis tasks are performed based on the one or more input medical images using the one or more selected machine learning based models. A guided review of the one or more input medical images is generated based on results of the one or more medical imaging analysis tasks. The guided review of the one or more input medical images is output.