Elena Provenzano shares results of the multi-center DeBORAH Study on diagnostic accuracy, and cost and efficiency savings, arising from implementation of an AI algorithm in breast core biopsy reporting, presented at ECP 2026.
Elena Provenzano is Lead Breast Histopathologist at Addenbrookes Hospital and Pathology Professional Clinical Advisor for National Health Service (NHS) BSP East of England.
The following transcript has been edited for clarity.
Hello, my name's Dr. Elena Provenzano. I'm lead breast pathologist at Addenbrooke's Hospital and principal investigator for the DeBORAH study, a multicenter study within the UK National Health Service (NHS) looking at diagnostic accuracy, and cost and efficiency savings, through implementation of an AI algorithm in breast core biopsy reporting.
The study involved five centers, including district general and university teaching hospitals. The first phase of the study was a validation phase using the AI as a second read following pathologist diagnosis. This demonstrated very high diagnostic accuracy, with greater than 99 percent sensitivity and specificity for overall cancer and invasive cancer diagnosis, and more than 96 percent sensitivity for diagnosis of ductal carcinoma in situ (DCIS).
In the second phase of the study, the AI was introduced as a first read as part of the diagnostic core biopsy pathway. This was pre-ordering of receptors for cases flagged as high likelihood of invasive cancer and assisting the pathologist in making the primary diagnosis.
Pathologists reported increased confidence in reporting with the assistance of the AI, and in two of the centers, there was an up to 17 percent reduction in ordering of immunohistochemistry, with one center showing a 72 percent reduction in ordering of immunohistochemistry in benign cases.
In four of the five centers, there was a reduction in cost ranging from £1.80 to £17.70 per patient. There was also a reduction in turnaround times of up to 16 to 17 hours.
A user satisfaction survey at the end of the study showed that the pathologists found the AI algorithm easy to use. It increased confidence, and they felt that they were working more efficiently and would recommend the AI to others.
So overall, in the NHS setting, the DeBORAH study demonstrated high diagnostic accuracy with reductions in cost and efficiency saving in breast core biopsy reporting. This was particularly noted in the district general hospital setting.
Newsletters
Receive the latest pathologist news, personalities, education, and career development – weekly to your inbox.
