For two years – between 2013 and 2015 – 737 patients were submitted to breast cancer screenings, 496 for prostate cancer and 348 for breast cancer, taking into consideration aspects such as age, weight, or family history.

In Portugal, about 4,500 new cases of breast cancer are detected annually, which result in the death of about 1,500 women. In the case of prostate cancer, the statistics are not very different, as about 4,000 cases are detected every year.

The use of computer technologies in medicine is increasing substantially, particularly in cancer. In fact, there is already the World Cancer Research Day, celebrated on 18 November.

The goal with ExpertBayes  was to combine medical knowledge with Machine Learning techniques in graphical models in order to estimate with more precision the risk of breast or prostate cancer after the biopsy. The idea is to reduce biopsy sampling errors.

This software is prepared to work with any type of data, and not necessarily those associated with cancer. However, in this specific case the prostate cancer data used have been provided by Carnegie Mellon University’s (CMU) Statistics Departments, and the breast cancer data have been provided by Elizabeth Burnside, breast cancer specialist at the hospital of the University of Wisconsin-Madison (USA) and co-author of the paper on the ExpertBayes that has now been published.

“The healthcare professionals have accepted the software really well because if there is any doubt during the diagnosis, the ExpertBayes helps them make the best decision. Part of this software is now available on Dataias and will be commercialised by the company NLPC Lda as part of a project by IAPMEI and the company NLPC, www.dataias.com”, explains Ezilda Almeida, one of the researchers at INESC TEC’s Centre for Research in Advanced Computing Systems (CRACS) responsible for developing this technology.

The software was developed as part of project ABLE (Advice Based Learning), led by Inês Dutra, researcher at INESC TEC and professor at the Faculty of Science of the University of Porto (FCUP). The project received over  88 thousand euros from the Portuguese Foundation for Science and Technology (FCT).

The researchers mentioned in this news piece are associated with the following institutions: INESC TEC and FCUP.