Notes to Editors
Looking ahead: Using Artificial Intelligence (AI) to improve accuracy and efficiency in breast screening
There is a radiologist supply shortfall in Scotland, compounded by long lead times for training both radiologists and advanced practitioners.
Currently, images are reported by two trained image readers, with specific training in mammography, who are required to meet performance criteria of the SBSP for both detection and recall rate.
Given the repetitive and high-volume nature of mammographic reporting, there is the potential for AI to maintain or increase cancer detection rates without increasing recall rates.
At this time, AI has not been approved for use within the national breast screening programmes. Within the SBSP, a research study is underway that will inform the UK National Screening Committee decision on whether AI can be adopted, but the timeframe for this is still unclear.
Critically, Philips’ technology enables health services to take advantage of AI screening, which can add a layer of qualification to the image screening process, reduce variation in analysis and provide radiologists with an essential second opinion. The net result may be increased capacity to see, screen and provide results to patients, whilst helping the NHS to tackle the significant backlogs created by COVID-19.
Dr Gerald Lip said:
“This will ultimately benefit women by providing faster results turnaround, relieve pressure on the workforce and reduce inequality providing a benefit to the entire population from the urban Centres to rural regions due to the national PACS solution providing a common interface for the algorithm.”