Generic integration of content-based image retrieval in computer-aided diagnosis

Area of Science:

• Medical Imaging Informatics
• Radiology
• Computer-Aided Diagnosis

Background:

• Content-based image retrieval (CBIR) shows promise for computer-aided diagnosis (CAD) but faces integration challenges in clinical practice.
• Interoperability issues between CBIR systems and clinical information systems hinder widespread adoption in radiology.
• Current Picture Archiving and Communication Systems (PACS) often exhibit deficiencies in Integrating the Healthcare Enterprise (IHE) compliance.

Purpose of the Study:

• To develop a novel integration scheme for seamlessly incorporating CBIR systems into radiological workflows.
• To address the integration gap by adapting existing IHE profiles for CBIR in CAD applications.
• To create a sustainable and reliable solution that fosters the use of CBIR in computer-aided diagnosis.

Main Methods:

• An intermediate integration scheme was developed, leveraging the IHE Post-Processing Workflow (PWF) integration profile.
• The scheme was adapted to the specific needs of CBIR within the context of CAD.
• The Image Retrieval in Medical Applications (IRMA) framework was utilized to exemplify the integration scheme, resulting in the IRMAcon application.

Main Results:

• The developed IRMAcon scheme provides a generic, convenient, and reliable method for integrating CBIR systems.
• The scheme is designed to be compatible with the evolving IHE compliance of PACS.
• Successful exemplification using the IRMA framework demonstrates the practical applicability of the integration scheme.

Conclusions:

• The IRMAcon scheme offers a sustainable solution for integrating CBIR into clinical systems and workflows.
• This approach effectively fosters the adoption and utilization of CBIR in computer-aided diagnosis.
• By addressing interoperability challenges, the scheme promotes the advancement of radiological routine with advanced diagnostic tools.