Radiology is becoming the bottleneck in oncological care, delaying treatment decisions. In the last twenty years, the number of thoraxabdomen CT scans of oncology patients increased over five-fold. Scans are larger, made more frequent, and oncologists request more precise and quantitative analysis to provide optimal care. OncoRead aims to address this bottleneck with a comprehensive AI-driven environment that allows radiologists to report more accurately and more efficiently. Our AI technology detects subtle changes, provides precise assessments of lesion progression, and ensures no new lesions are overlooked.
OncoRead’s founder/investor is Bram van Ginneken. He is professor of medical image analysis, built the most widely used autonomous AI product in radiology (CAD4TB), developed the fastest reading environment for CT lung cancer screening (Veolity) and founded the scale-up Thirona. Co-founder and chief medical officer is Mathias Prokop, CT innovator and radiologist, head of two radiology departments in The Netherlands, and former chair of the Dutch Society of Radiology. We are backed by the largest research groups in medical image analysis in the Netherlands (DIAG, Radboudumc) and Europe (Fraunhofer MEVIS, Bremen, Germany). We license IP from DIAG and MEVIS, based on research performed since 2016.
Our unique edge lies in our partnership with a network of Dutch radiology departments, providing data and feedback, allowing continuous improvements of our AI engine and web-based front-end. We secured a contract with a notified body that focuses exclusively on medical device software, allowing us to obtain CE certification (class IIb) in two years. We have compiled a dataset of over 200,000 annotated CT thoraxabdomen studies and already developed >40 algorithms. This will be extended to >100 to cover >80% of the findings included in radiology reports and reduce reporting time by more than 50%.