JAMIA: How to transfer content-based image retrieval into practice
Adapting a DICOM structured reporting template to content-based image retrieval (CBIR) could help integrate CBIR into PACS workflow and clinical practice, according to an article published in the June issue of the Journal of the American Medical Informatics Association.

CBIR brings inherent advantages to image management by leveraging image content and visual information for retrieval. In contrast, archives employ DICOM standards, which use alphanumerical annotations, which are prone to manual errors and may result in failed queries, to query images in the database.

Computer-assisted detection (CAD) may provide an early application of CBIR, with CBIR automatically identifying relevant prior studies, for the radiologist, explained Petra Welter, of the department of medical informatics at Aachen University of Technology in Aachen, Germany, and colleagues.

According to Welter, one of the reasons CBIR has not yet been adopted into clinical practice is its lack of integration into PACS workflow. That is, current CBIR systems are stand-alone workstations.

The authors developed and tested a model to apply CBIR to CAD and proposed that the following items be included in the CBIR document:
  • CBIR system;
  • Image or region of interest of query;
  • Reference database;
  • Result images; and
  • Similarity scores.

“We suggest using the DICOM standard for reliable data integration,” wrote Welter, who built and tested a working system for CBIR and PACS integration.

The authors demonstrated the IRMACON SR viewer, a user interface that combines hanging protocols with data automatically pulled the RIS. The viewer includes an entry page with thumbnails of key images and a detail page with diagnostic content.

The authors tied their work to DICOM Supplement 118, which proposes an open, standardized application programming interface for a DICOM-based image processing and analysis system. Tying systems together can help further integrate CBIR and PACS, the researchers offered.

The system addresses the lack of integration between CBIR and CAD, the authors continued. “To obtain the predicted benefits for CAD, CBIR results must be encoded in a standardized format and integrated into PACS and HIS to meet radiologists’ workflow requirements. ...The proposed system architecture using DICOM SR documents may bridge this gap.”

Around the web

Eleven medical societies have signed on to a consensus statement aimed at standardizing imaging for suspected cardiovascular infections.

Kate Hanneman, MD, explains why many vendors and hospitals want to lower radiology's impact on the environment. "Taking steps to reduce the carbon footprint in healthcare isn’t just an opportunity," she said. "It’s also a responsibility."

Philips introduced a new CT system at ECR aimed at the rapidly growing cardiac CT market, incorporating numerous AI features to optimize workflow and image quality.

Trimed Popup
Trimed Popup