Archive
OPI projects

AI-augmented Radiology: Detection, reporting, and clinical decision-making in prostate cancer diagnosis

The National Information Processing Institute (OPI PIB) has received funding to develop an innovative platform that facilitates prostate cancer diagnosis.

Project data

Project start date2021-12-02
Project end date2025-06-01
Project cost7347082.5 PLN
Project amount of funding7347082.5 PLN

Lider

  • National Information Processing Institute

Subcontractor

  • Lower Silesian Oncology Center (scientific partner of OPI PIB)

The National Information Processing Institute (OPI PIB) has received funding to develop an innovative platform that facilitates prostate cancer diagnosis. The cutting-edge tool will rely on artificial intelligence (AI) to assist medical experts in implementing the most optimal therapies.

Prostate cancer has become a growing problem, both from the medical and social perspectives. Diagnostic methods, such as medical imaging and, in particular, multi-parametric magnetic resonance imaging (mpMRI), are becoming increasingly popular among clinicians. The analysis of mpMRI scans is a complex and multifaceted task; the quality of reporting of lesions and communication between radiologists and clinicians play key roles in implementing suitable diagnosis and treatment plans. With staff shortages among radiologists and the long training necessary to become an experienced specialist that is capable of properly evaluating mpMRI results, the use of modern technologies to enhance radiologists’ cognitive abilities and to facilitate specialists’ tasks is imperative.

The main goal of the project is to create a system that reports prostate cancer mpMRI scans; one that is developed on the basis of integration between standardised structural reports and trained machine learning (ML) methods that analyse mpMRI scans—a highly innovative concept. The results of the image analyses will be used to support radiologists in various aspects of their work. These will include: automatic measurement of the size and calculation of the volume of prostate glands; creation of visual cues on the locations of suspicious lesions; evaluation of the clinical relevance of identified lesions on the PI-RADS scale; automatic pre-filling of fields in template reports; and searching for similar cases.

The project’s results will be made available via an e-learning platform that will train its users how to structurally report prostate mpMRI scans according to PI-RADS using a dedicated reporting system. Educational resources will be provided by an integrated reference base of imaging data that contains an extensive set of thoroughly described cases and full medical documentation. Both the database and the tasks related to medical aspects will be developed alongside the Lower Silesian Oncology Center.

Subsidies

The AI-augmented radiology: Detection, reporting, and clinical decision-making in prostate cancer diagnosis project has been devised by experts at OPI PIB led by Dr Rafał Jóźwiak. The project will be subsidised as part of the Medical Image Recognition task (the First INFOSTRATEG competition, Strategic Programme for Research and Development – Advanced information, telecommunication and mechatronics  technologies).

The results of the First INFOSTRATEG competition have been published here by the National Centre for Research and Development.

The project is financed by the National Centre for Research and Development as part of the INFOSTRATEG I programme.