Related Experiment Video
Updated: Sep 13, 2025
05:33
Introduction of an Integrated Pathology Image Management, Artificial Intelligence, and Reporting System
Published on: July 11, 2025
251
Deep learning-based image classification for integrating pathology and radiology in AI-assisted medical imaging
Chenming Lu1, Jiayin Zhang2, Ren Liu3
1Shanghai General Hospital, Shanghai, China.
Scientific Reports
|July 27, 2025
Summary
This study introduces a novel AI framework, Domain-Informed Adaptive Network (DIANet), and Adaptive Clinical Workflow Integration (ACWI) to improve medical imaging diagnostics. The approach enhances accuracy and reliability in clinical settings by addressing multi-modal data and interpretability challenges.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Pathology and Radiology Integration
Background:
- Current AI in medical imaging faces challenges with multi-modal data, dataset imbalance, interpretability, and uncertainty quantification.
- These limitations impede the clinical deployment of AI systems, necessitating more robust and adaptable solutions.
- Integrating pathology and radiology is crucial for enhancing diagnostic accuracy and clinical workflows.
Purpose of the Study:
- To introduce a novel AI framework, the Domain-Informed Adaptive Network (DIANet), and an Adaptive Clinical Workflow Integration (ACWI) strategy.
- To address limitations in current AI for medical imaging, including multi-modal analysis, imbalanced datasets, and interpretability.
- To enhance diagnostic accuracy, segmentation precision, and reconstruction fidelity in medical imaging.
Main Methods:
- Developed DIANet utilizing multi-scale feature extraction, domain-specific priors, and Bayesian uncertainty modeling.
- Integrated adaptive learning mechanisms within DIANet to handle domain shifts and imbalanced datasets.
- Implemented ACWI strategy with explainable AI (XAI), uncertainty-aware decision support, and PACS-compatible modular integration.
Main Results:
- Demonstrated significant improvements in diagnostic accuracy across diverse medical imaging modalities.
- Achieved enhanced precision in medical image segmentation tasks.
- Validated improved reconstruction fidelity, showcasing the framework's effectiveness.
Conclusions:
- The proposed DIANet and ACWI framework effectively bridges the gap between AI innovation and clinical utility in medical imaging.
- The framework offers enhanced interpretability, robustness, and adaptability for real-world clinical applications.
- This integrated approach holds significant potential for advancing diagnostic capabilities and clinical workflows in pathology and radiology.
