– AI tool Mia successfully detects signs of breast cancer in mammograms
– Trained on a dataset of over 6,000 breast cancer cases, AI system accurately predicts cancer with 81.6% accuracy
– Researchers developing AI further to predict risk of side effects after treatment, planning clinical trial to validate AI risk prediction model
A successful AI tool called Mia was piloted in the UK alongside NHS clinicians, analyzing mammograms of over 10,000 women. The AI flagged symptoms of breast cancer in all affected patients, including 11 cases missed by human doctors. The tool was trained on a dataset of previous breast cancer cases and accurately predicted the presence of cancer 81.6% of the time.
Breast cancer is a common disease in women globally, with two million new cases reported annually. While survival rates have improved, patients still suffer from severe side effects like lymphoedema after treatment. Researchers are now working on enhancing the AI system to predict a patient’s risk of such side effects up to three years post-treatment, enabling personalized care for high-risk patients.
The research team plans to conduct a clinical trial called Pre-Act, enrolling 780 breast cancer patients to validate the AI risk prediction model over a two-year follow-up period. The ultimate goal is to develop an AI system that can comprehensively evaluate a patient’s prognosis and determine their treatment needs. The innovative use of AI in healthcare could revolutionize cancer diagnosis and treatment, enhancing patient outcomes and quality of care.
To learn more about AI and big data in healthcare, industry leaders can attend the AI & Big Data Expo in Amsterdam, California, and London. This event, co-located with other leading tech expos, provides a comprehensive overview of the latest advancements in AI technology and how they are shaping the future of healthcare. Attendees can explore upcoming events and webinars powered by TechForge for more insights into emerging trends in enterprise technology.