Radiology Explorer is a fast and efficient 3D image and text search engine. For radiologists it reduces the time to find similar images from ~10 minutes to less than 1 minute, compared to the current text-only based search procedure. Radiologists simply mark a region of interest in an image and within 5 seconds we compare it to all data stored in the hospital and return the visually most similar ones.
Our product is the result of a 4 year european research project and was developed by the Computational Imaging Research Laboratory (CIR) in close collaboration with radiologists at the Department of Biomedical Imaging and Image-guided Therapy at the Medical University of Vienna, where ~250,000 recordings (e.g. CT, MRI and x-ray) are performed each year.
The system combines state of the art machine learning methods applied to medical imaging data with clinical knowledge of radiologists and a cloud based architecture. While its primary use is clinical routine, its core technology enabl...
Radiology Explorer is a fast and efficient 3D image and text search engine. For radiologists it reduces the time to find similar images from ~10 minutes to less than 1 minute, compared to the current text-only based search procedure. Radiologists simply mark a region of interest in an image and within 5 seconds we compare it to all data stored in the hospital and return the visually most similar ones.
Our product is the result of a 4 year european research project and was developed by the Computational Imaging Research Laboratory (CIR) in close collaboration with radiologists at the Department of Biomedical Imaging and Image-guided Therapy at the Medical University of Vienna, where ~250,000 recordings (e.g. CT, MRI and x-ray) are performed each year.
The system combines state of the art machine learning methods applied to medical imaging data with clinical knowledge of radiologists and a cloud based architecture. While its primary use is clinical routine, its core technology enables researchers to discover novel imaging biomarkers, and exploration of big datasets stored in hospitals for teaching purposes. We have the support of the head of the Department of Biomedical Imaging and Image-guided Therapy (Univ. Klinik für Radiologie und Nuklearmedizin) at the AKH, where we will also install our first research and teaching system in the beginning of next year, followed by a commercial version of the system. After two rounds of successful user tests finished already, this will establish our product in the medical domain
Our core image based search system is the next step for picture archiving and communication systems (PACS). It goes beyond image management and enables learning from the vast amount of knowledge encoded in ever growing medical image data stored in hospitals.
From our worldwide network to clinicians and researchers we know that there is currently no such system on the market. We are the first go-to partner for unlocking the potential of medical image data and CIR already has research collaborations with industry partners in this direction. We will licence our core engine to PACS vendors giving them a competitive advantage and simultaneously distributing our system to the heart of the clinics. Our long-term goal is to establish a key position for point of care use of large-scale medical image data analysis.
More information
