Bridging the text-image gap: a decision support tool for real-time PACS browsing.

Area of Science:

• Medical Imaging
• Ontology Engineering
• Knowledge Management

Background:

• Clinical workflows often face challenges integrating medical imaging data with external knowledge sources.
• Accessing relevant information for Magnetic Resonance (MR) image interpretation can be time-consuming and inefficient.

Purpose of the Study:

• To introduce an ontology-based technology that bridges the gap between MR images and external knowledge sources.
• To enhance the accuracy and efficiency of clinical case assessment by facilitating access to pertinent information.

Main Methods:

• Development of an ontology-based system for linking MR images to knowledge bases.
• Implementation of a user interface allowing manual selection of regions of interest (ROIs) on MR images.
• Automatic inference of anatomical structures from user-selected ROIs to query external knowledge sources.

Main Results:

• The technology demonstrated high recall (>95%) in accurately inferring the intended brain region from manual selections in a user study.
• Evaluation involved three neuroradiologists, indicating practical applicability in the neurodomain.
• The system effectively connects clinical image data with external knowledge repositories.

Conclusions:

• The proposed ontology-based technology successfully bridges the gap between MR imaging and knowledge sources.
• This approach enhances diagnostic certainty, accuracy, and efficiency in clinical practice.
• The system improves the user experience for accessing external knowledge relevant to MR image interpretation.