Royal Free to study AI tool for breast cancer screening

Royal Free London Hospital NHS Foundation Trust is to carry out a retrospective study to determine if artificial intelligence could potentially detect breast cancer more effectively than current methods.

As part of NHSX’s AI Award, the innovation and intelligent automation team at the trust is teaming up with North London Breast Screening Service (which is run by the the trust at Edgware Hospital) to assess Kheiron Medical Technologies’ artificial intelligence (AI) tool, Mia. It will look at AI’s ability to detect breast cancer by reviewing mammograms previously run at North London Breast Screening.

If the study proves successful, then the AI tool could be explored further, as a potentially faster and more accurate method of diagnosing breast cancer.

Current screening methods see mammograms reviewed by two human ‘readers’. Although using two people to assess a mammogram improves cancer detection rates, it requires more resources than screening by a single person. If Mia is effective, eventually the need for a second ‘reader’ could be redundant, which would help tackle the lack of resources, backlogs to the screening service and staff shortages.

Dr William Teh, clinical site lead at the trust, said: “The study will provide evidence to help assess whether the use of AI in this instance could be a viable option. It’s an opportunity to analyse and assess the differences between human and machine made decisions and to learn whether AI has the potential power to positively transform clinical practice in real-world screening populations.”

The NLBSS is one of 14 sites taking part in the study which will involve Mia being used to analyse the mammogram image and then recommending either no further action is required, or that further investigation is needed. Mia will be studying mammograms that have already been reviewed by two ‘readers’, and at the end of the study the diagnoses made by Mia and the ‘readers’ will be compared.

The more opportunity Mia has for analysis, the better it gets at spotting cancer, so the process will help improve the tool’s ability to detect cancerous or pre-cancerous cells. It will also give scientists the opportunity to see how AI performs compared to traditional screening methods.

The technology has previously been tested at a smaller scale, but the study is hoping to support the wider roll out of the tech by proving that Mia can be used for diverse and large screening populations.

AI is also currently being explored at King’s College Hospital NHS Foundation Trust, where it is analysing scans and data to help with diagnostics and treatment plans for the trust’s stroke patients.