AI helps detect kidney cancer faster

Researchers at the University of Tartu Institute of Computer Science, Junior Research Fellow of Artificial Intelligence Joonas Ariva and Associate Professor in Artificial Intelligence Dmytro Fishman, together with radiologists from Tartu University Hospital and engineers at Better Medicine, have developed an AI-based tool that accelerates the detection of kidney cancer. Its effectiveness was validated in a study published in Nature Communications Medicine.

Diagnosing kidney cancer relies on computed tomography (CT) scans and the use of contrast agents. These examinations provide radiologists with a detailed overview of kidney structure and potential abnormalities. It is important to detect kidney cancer also on imaging studies that were not performed specifically to search for a tumour, but for other reasons, such as trauma or abdominal pain. 

One of the major global challenges in radiology is the shortage of radiologists, while the number of imaging studies continues to increase year by year. This further increases the workload of an already overstretched workforce and makes the role of additional support tools in image interpretation increasingly important. In this context, AI-based tools can provide valuable support to radiologists.

To support clinicians in interpreting complex medical images, computer scientists at the University of Tartu, in collaboration with Better Medicine, have developed a machine-learning-based solution called BMVision. It analyses CT images and helps radiologists detect both malignant and benign lesions more quickly and reliably.

The effectiveness of BMVision was evaluated in a retrospective study conducted at Tartu University Hospital and published in Nature Communications Medicine. Six radiologists reviewed 200 CT scans, each in two ways: with and without AI assistance. This resulted in 2,400 individual readings, which were compared across several clinical indicators, including diagnostic sensitivity, accuracy of tumour measurement, reporting speed, and inter-radiologist agreement.

The results were striking. Using AI reduced the time needed to identify, measure, and report malignant lesions by roughly one third.

The study confirms that AI does not replace the radiologist but serves as a reliable assistant and a second set of eyes. Such tools enable doctors to focus more time on cases that need it most, while giving patients a better chance of an early diagnosis.

“This study adds to the growing body of evidence that modern AI tools developed in research labs can make a real impact in clinical practice and support doctors in their daily work. We are very encouraged by these results, which show that AI research in medicine is not only meaningful, but it can truly be used for good,” commented Associate Professor in Artificial Intelligence and co-founder of Better Medicine Dmytro Fishman.

According to Professor of Radiology at Tartu University Hospital, Dr Pilvi Ilves, the introduction of the AI-based tool may improve diagnostic quality and enable earlier detection of kidney cancer. “While the solution has so far been used at Tartu University Hospital only for research purposes, it is now being integrated into the clinical workflow. In the future, all abdominal CT scans performed at our hospital will be processed through BMVision,” Ilves said.

Better Medicine obtained a CE marking for BMVision, confirming that the product meets the environmental, health, and safety standards of the European Economic Area. This makes BMVision the first AI tool available on the market that helps detect kidney cancer early and assess it more accurately.