AI set to transform Breast Cancer Diagnosis

UK Clinical Trial Sets the Stage for AI-Powered Breast Cancer Diagnosis

Aanchal Khanna

Pravin Gandhi College of Law

On World Cancer Day, the National Health Service, UK, which is a public healthcare service, announced a clinical trial that would use Artificial Intelligence for early diagnosis and detection of breast cancer, which is the most common cancer in women as 55,000 are diagnosed every year. There would be around 700,000 women in the UK who would be participating in this clinical trial. The trial is launched in light of the UK government’s National Cancer Plan, a part of its 10-Year Health Plan, which seeks to maximize its impact through up-to-date technology and innovations. The £11 million trial is set to begin in April and it will test five AI platforms across 30 sites.

A mammogram, which is a low-dose x-ray system that examines the breasts, is done to detect breast cancer and is conducted in most cases for early detection of breast cancer. It requires the two radiologists to observe the x-ray and diagnose whether there is breast cancer or not. According to the National Library of Medicine, Radiologists miss about 10-20% of breast cancers in mammograms, 80% called again for more views have normal outcomes, with 40% of biopsied lesions being benign[1].

According to the American Cancer Society, half of all women who get screenings over 10 years have a false positive result.

To ensure proper and accurate diagnosis, AI technologies are being used as an assist to radiologists, due to which there would only be a requirement of one radiologist for the diagnosis instead of two. AI technologies and software can register subtle changes in tissues that radiologists can miss. It can read chest X-rays or CT scans and spot lung nodules and tumors. It can identify rare gene mutations in hereditary cancers. It can also predict patient responses to specific therapies, optimizing treatment and reducing toxicity. All in all, AI technologies are efficient and highly accurate, thus the use of it in breast cancer detection allows for early diagnosis and treatment.

There was a study done at the Capio Sankt Göran Hospital in Stockholm, Sweden called Screen Trust CAD in which for the breast cancer diagnosis reading, there was a comparison between one radiologist with AI and two radiologists. The result was that the combination of the radiologist and AI detected more cases of breast cancer than the combination of the two radiologists.[2] AI can observe what humans are not able to and it is thus important that both work together to achieve accurate and the greatest results.

This clinical trial is therefore a significant and crucial step towards enhancing breast cancer diagnosis and detection, as it would make it faster and more accurate for women to start treatment at the earliest. It is called an Early Detection using Information Technology in Health (EDITH) trial as the successful use of AI could reduce the pressure and burden on the radiologists by making the process seamless. The radiologists can then use their time to save more lives.

Wes Streeting, the Health Secretary of the UK, while talking about the trial said that it was the beginning of improving cancer treatment and care as many patients face long queues and delayed diagnoses in the care they require. Lucy Chappell, the Department of Health and Social Care’s chief scientific adviser, described the study as a landmark trial in leveraging technology to enhance patient care.

Dr Katharine Halliday, President of the Royal College of Radiologists, said that it was amazing to see the government of the UK launching a trial AI had immense potential to ensure safe and quick diagnostics and improve cancer care and it was trial but warned that there were also risks involved, due to which clinical oversight would be necessary. She said that even though the trial would reduce the pressure from the radiologists, it is important to note that there is a need to build the capacity of the radiology department as the trial is a study that will take time with a projected 40 percent shortfall in consultant radiologists by 2028.[3]

In conclusion, this clinical trial is a crucial step as it will transform breast cancer diagnosis as it would be more accurate, quicker and efficient and would reduce the possibility of false negatives or positives. The clinical trial highlights the potential to transform breast cancer diagnosis and any kind of cancer diagnosis, which would improve patient treatment as in many cases early treatment would be possible due to early diagnosis by AI.

References

[1] https://pmc.ncbi.nlm.nih.gov/articles/PMC5980911/#:~:text=However%2C%20high%20error%20rates%20in,of%20biopsied%20lesions%20being%20benign.

[2] https://www.thelancet.com/journals/landig/article/PIIS2589-7500(23)00153-X/fulltext 

[3] https://www.independent.co.uk/news/health/ai-breast-cancer-mammogram-b2691244.html