Researchers at Lund University in Sweden have developed methods to simulate how breast tumours grow and how breast tissue changes over time. By digitally replicating both the anatomy of the breast and the development of tumours, new, advanced imaging technology can be tested more quickly and safely than is currently possible - without exposing patients to radiation or subjecting them to further examinations.
Breast cancer is the most common form of cancer among women worldwide. Early detection using breast imaging, such as mammography, is crucial for saving lives. At the same time, diagnostic imaging is developing rapidly, with increasingly advanced analysis and the use of artificial intelligence. For such technologies to be safely introduced into healthcare, they must be able to be evaluated - preferably without placing an additional burden on patients.
In a new thesis from Lund University, methods have been developed for virtual diagnostic imaging studies - computer-based models capable of simulating both healthy and diseased breasts. The aim is to simulate anatomical changes in the breast in order to better understand how breast cancer develops and appears in X-ray images.
The research in question at Lund University has analysed real patient data to improve our understanding of how tumours grow over time. The results have then been used to create realistic, computer-simulated tumours.
"The next step was to create a virtual environment in which the entire breast can be simulated and changes in tissue and tumours can be monitored over time," says Hanna Tomic, doctoral student and a member of the Diagnostic Radiology research group at Lund University.
This makes it possible to create digital 'patients' who develop breast cancer in different ways, and then test how effectively various imaging techniques detect these changes. The researchers have also shown that the simulated tumours can be combined with real mammogram images in what is known as hybrid studies. This makes it possible to investigate the conditions for detecting breast cancer in different groups of women.
The results show that the models correlate well with real patient data and that the simulated tumours resemble those seen in clinical practice. In a virtual study, the researchers found that breast cancer is more difficult to detect in women with dense breast tissue, a finding previously observed in clinical trials.
"The findings of the thesis show that virtual imaging studies can accurately simulate how breast cancer develops and appears on X-rays. This paves the way for faster and more cost-effective methods of evaluating new imaging techniques - and for improving the chances of detecting breast cancer at an early stage," says Predrag Bakic, associate professor and researcher in diagnostic radiology at Lund University.
