Inside a dark room at Bács-Kiskun County Hospital outside Budapest, Dr. Éva Ambrózay, a seasoned radiologist, made an important discovery. Utilising artificial intelligence AI technology, she identified potential signs of breast cancer that two human radiologists had missed. This breakthrough represents a significant advancement in breast cancer screening and highlights the transformative potential of AI in improving public health. Hungary, with its robust breast cancer screening program, serves as a major testing ground for this technology, which is now demonstrating comparable accuracy to human radiologists.
The promise of AI in breast cancer screening
Breast cancer screening has historically relied on the expertise of radiologists to interpret mammograms accurately. However, the complexities of identifying subtle signs of cancer within mammogram images can pose challenges, leading to potential oversight. AI offers a solution by leveraging machine learning algorithms to detect patterns and anomalies that may elude human eyes. Hungary, alongside the United States, the United Kingdom, and the European Union, is at the forefront of testing and implementing AI systems for breast cancer screening.
Overcoming challenges and resistance
Although AI holds immense promise, widespread adoption still faces several obstacles. Extensive clinical trials are necessary to ensure the technology’s reliability as an automated second or third reader of mammograms. It must demonstrate accuracy across diverse age groups, ethnicities, and body types. Additionally, the AI tools need to effectively identify complex forms of breast cancer while minimising false-positive results. These challenges underscore the importance of continued collaboration between AI developers and medical professionals.
The human and AI partnership
Debates have emerged regarding the potential replacement of human radiologists by AI technologies. However, industry experts emphasise that AI should be seen as a tool to enhance radiologists’ capabilities, rather than replacing them entirely. Regulatory scrutiny and resistance from some doctors and healthcare institutions indicate a preference for maintaining the human and AI partnership in clinical settings. Patients’ trust and the effectiveness of the technology rely on the integration of AI with the expertise of trained medical professionals.
A lifesaving potential for breast cancer detection
Dr. László Tabár, a renowned mammography educator in Europe, emphasises that AI could ultimately save lives. Extensive performance reviews of various AI vendors’ breast cancer screening technologies have convinced Dr. Tabár of their efficacy. Early trials have demonstrated AI’s ability to identify developing cancers that human radiologists may have missed, offering potentially life-saving interventions. These encouraging results provide hope for improved early detection and better patient outcomes.
The journey of artificial intelligence in breast cancer screening
The roots of AI in detecting illness and disease can be traced back to image recognition systems developed by pioneers like Geoff Hinton and his colleagues at the University of Toronto. Neural networks, modelled after the human brain, form the core technology behind these systems. Inspired by this success, AI evangelists sought to apply similar techniques to breast cancer detection in mammograms. The sheer number of breast cancer diagnoses and deaths worldwide underscores the urgent need for accurate and efficient screening methods.
How AI is empowering radiologists
The founders of Kheiron Medical Technologies, Peter Kecskemethy and Tobias Rijken, recognized the need to support radiologists in their crucial work. By leveraging AI, they developed systems capable of analysing millions of historical mammograms to identify cancerous growths accurately. Kheiron’s AI technology creates a mathematical representation of normal mammograms and compares it to each new image, enabling it to detect abnormalities imperceptible to the human eye. Results from trials conducted in Hungary demonstrated the AI’s ability to match human radiologists’ performance and increase cancer detection rates.
Validation in Hungary and beyond
MaMMa Klinika, a Budapest-based clinic, became the first site to use Kheiron’s AI technology on patients. By supplementing the expertise of two radiologists, the AI system has successfully identified cancers overlooked by human interpretation. MaMMa Klinika’s experiences, along with positive results from other Hungarian clinics, have established the technology as a significant breakthrough. Furthermore, Scotland’s National Health Service, England’s National Health Service, Oulu University Hospital in Finland, and medical screenings in Oman are incorporating AI technology to enhance breast cancer detection.
Embracing the future
While artificial intelligence technology holds enormous potential to address the 20 percent of breast cancers missed during mammography screenings, the expertise of radiologists remains vital. Dr. Constance Lehman, a prominent radiology professor, emphasises the importance of doctors maintaining an open mind regarding the integration of AI into healthcare practices. Computers excel at specific tasks, and their collaboration with human experts can lead to improved accuracy and efficiency in breast cancer screening.
Artificial intelligence is revolutionising breast cancer screening, bridging the gap between radiologists’ expertise and the potential for overlooked abnormalities. Early results from Hungary and other testing grounds suggest that AI systems can match or surpass human radiologists’ performance, potentially saving lives through early detection. Collaboration between AI developers and medical professionals are key to further refining the technology, addressing challenges, and ensuring its integration as a valuable tool in healthcare. As the journey of AI in breast cancer screening continues, the future holds promise for improved outcomes, reduced oversight, and enhanced patient care.