Use of Artificial Intelligence in breast cancer research

Breast cancer screening

AI can be used in the assessment of mammography images. In fact, in Denmark, it has already been implemented in population screening. In addition to assessment by radiologists, a third evaluation by AI is an additional asset. This not only reduces the risk of erroneous assessments, but also immediately reduces the workload of radiologists. 

Ongoing clinical studies

More and more clinical studies are using AI in evaluating radiological images of breast tumors. Comparisons with radiologist-researcher evaluations show that the accuracy of AI results is good, making it possible for future studies to be faster and more efficient.

Support for breast cancer diagnosis

AI models can already identify suspicious spots on a mammogram and assign a probability score for breast cancer to them.Combined with a radiologist's judgment, fewer breast cancers are missed this way.

Decision

Despite promising results, challenges remain. AI models are mainly tested on radiological images that radiologists have already selected for suspicious images. Whether these tools alone will be sufficient to assess screening mammography, for example, will require further research.