AI assisted breast cancer diagnosis

Our funding allowed the pathology team at Betsi Cadwaladr University Health Board to test the use of AI software (IBEX Galen AI platform) in performing the first diagnostic step in cases of breast cancer, looking at the effectiveness and impact of integrating this step into the pathway. This allowed the health board to build on previous work which saw them test the same software in the diagnosis of prostate cancer.

In our project, AI software was introduced to the beginning of the specimen journey to provide an initial evaluation of the scanned breast biopsies. The software identified areas of malignancy, highlighting them for clinical staff to examine, and triggering a request for a receptor profile of the biopsy to be undertaken before the case was allocated to a pathologist.

Ordinarily, the request for the receptor profile is submitted by the pathologist after first examining the specimen; they then undertake a final evaluation of the biopsy once the receptor profile has been received. By integrating software into the pathway to carry out the initial evaluation and order the profile, the specimen journey can be reduced, enabling results to reach patients, and clinical decisions to be made more quickly.

Ultimately, the reporting remains firmly in the hands of the pathologist, however, this pathway innovation can reduce the time from biopsy to diagnosis and treatment, and it adds another level of scrutiny where AI can support pathologists to more easily identify cancers.

Reflecting on the work, Dr Muhammad Aslam, Consultant Pathologist and Clinical Lead in North Wales, said

"Early detection saves lives and AI in Betsi Cadwaladr is improving prostate and breast cancer diagnoses. This is just the start and it will lead to quicker diagnoses and more accurate predictions of prognoses for patients with various cancers".