Hong Kong witnesses an average of 15 women being diagnosed with breast cancer daily, making it the most prevalent cancer among women in the city and the third leading cause of cancer-related deaths. In 2021, breast cancer constituted 28.5% of all new cancer cases in Hong Kong among women. Globally, breast cancer tops the charts as the most frequently diagnosed cancer and the leading cause of cancer death in females, with an estimated 2.3 million women diagnosed in 2020 and 685,000 succumbing to the disease that same year.
While mammography screening has been the linchpin for early detection since the 1980s, variations in radiologists’ diagnostic accuracy and a global shortage of breast radiologists have led to challenges. Hong Kong is now exploring the integration of artificial intelligence (AI) to address these issues.
The Hospital Authority in Hong Kong has already implemented AI for triage and detecting abnormal findings in chest X-rays. Radiologists in Hong Kong are now testing and using AI software specifically designed for breast imaging. This move follows similar initiatives in Sweden, Hungary, and the United States, among other locations, where deep-learning AI software has shown potential to enhance breast cancer detection rates and streamline mammogram interpretation.
Dr. Alta Lai Yee-tak, a consultant radiologist at Pamela Youde Nethersole Eastern Hospital and honorary secretary of the Hong Kong College of Radiologists, emphasizes the excitement within the radiology field about the potential of AI. Unlike traditional computer-aided detection (CAD) used in mammography for over two decades, AI has the ability to learn from vast datasets, potentially identifying patterns beyond human capabilities.
Radiologists Dr. Julian Fong Chun-yan and Dr. Lui Chun-ying of Hong Kong Women’s Imaging (HKWI) have been testing the Lunit Insight MMG, an AI solution for mammography, since 2022. They have integrated the FDA-approved software into their workflow, processing over 10,000 mammograms. HKWI is the first and only breast screening center in Hong Kong formally using AI on a daily basis, distinguishing it from public hospitals where AI mammography solutions are still under trial.
Dr. Fong notes that Insight MMG’s suitability for Hong Kong is attributed to its development with a significant dataset from South Korean women. Considering the differences in breast composition between Asians and Caucasians, particularly the higher proportion of dense breast tissue in Asian women, having AI designed with Asian-specific samples is crucial.
Breast surgeon Dr. Hung Wai-ka acknowledges AI as a supplementary tool in breast cancer screening, potentially enabling the detection of more early-stage cancers. Early detection allows for less invasive treatments, reducing surgery, lymph node procedures, and overall treatment costs.
Despite the promise of AI, Dr. Fong emphasizes that AI software is not flawless, and experienced breast radiologists remain essential for accurate interpretation. Collaboration between human expertise and AI is deemed complementary, with AI serving as a supplementary alert system for radiologists.
While cautious optimism surrounds the integration of AI in breast cancer screening, challenges such as biases in training data and the need for thorough follow-ups to confirm AI diagnoses’ accuracy must be addressed. The medical community in Hong Kong anticipates continuous improvements in AI technology, providing enhanced healthcare outcomes in collaboration with human expertise.