The fight against breast cancer is far from over, but with AI as an ally, we are more prepared to confront this challenge directly.
This is the perspective of renowned surgeon and futurist Professor Shafi Ahmed, who recently shared his insights on the rapidly evolving role of AI in addressing a significant global health challenge.
Breast cancer is the leading cause of cancer-related deaths among women. In 2022 alone, there were 2.3 million diagnoses and 670,000 fatalities, making early detection and personalised treatment more crucial than ever.
Despite significant advances in screening, challenges such as missed diagnoses and disparities in healthcare access continue to persist.
AI is revolutionising breast cancer care by enhancing diagnostic accuracy and facilitating more targeted treatment strategies.
It is emerging as a transformative force in breast cancer detection, enhancing both efficiency and precision in screening. Although traditional mammography is widely used, it has its limitations – studies suggest that up to 20% of breast cancers may be overlooked.
AI-driven solutions, such as those discussed by Professor Shafi Ahmed in his latest AI Horizons newsletter, are transforming this landscape by minimising false positives and facilitating earlier detection.
Among the most notable advancements is Mia from Kheiron Medical Technologies. This AI-powered platform, trained on millions of mammograms, serves as an independent second reader during screenings.
Solutions like Mia Reader help radiologists make critical decisions, while Mia Triage points out suspicious cases for more precise oversight.
By automating up to 25% of the reading workflow, Mia tackles workforce shortages, alleviates radiologist fatigue, and enhances cancer detection rates. Studies from Hungary and the UK have demonstrated its potential, showing that Mia increases detection rates by 7-12% and significantly lightens radiologists’ workloads.
Research continues to highlight AI’s potential in breast cancer care. A recent study analysed Germany’s mammography screening programme and revealed a 17.6% increase in cancer detection rates in AI-supported screenings compared to traditional methods, all without an increase in recall rates.
Likewise, the PRAIM study underscored the enhancements in diagnostic accuracy and efficiency brought about by AI, alleviating the workload for healthcare professionals.
Further extending the boundaries of AI in oncology, researchers at Harvard Medical School developed CHIEF (Clinical Histopathology Imaging Evaluation Foundation), a versatile AI system that can diagnose 19 types of cancer, including breast cancer.
CHIEF interprets pathology images with near-human accuracy, detecting cancer cells, predicting tumour profiles, and even forecasting patient survival rates. In trials, CHIEF achieved 94% accuracy in cancer detection, surpassing existing AI models.
Meanwhile, researchers at MIT have unveiled Mirai, a cutting-edge AI model that can identify breast cancer up to five years prior to a clinical diagnosis. By analysing mammography images using deep learning algorithms, Mirai detects precancerous changes in tissue, offering high-risk patients access to earlier and more effective interventions.
While AI’s integration into breast cancer screening is still evolving, its potential is undeniable. Shafi said healthcare providers must prioritise regulatory approvals, clinician training and patient trust to ensure widespread adoption. Ethical considerations, such as algorithm transparency and data bias, must also be addressed to guarantee fair and equitable access to AI-powered care, he added.
While the convergence of artificial intelligence and breast cancer detection presents transformative possibilities, he is quick to point out that AI is not here to replace radiologists but to augment their expertise.
‘Navigating this landscape requires ensuring these advancements are accessible, equitable and implemented with the utmost care.’
