Artificial Intelligence (AI) has brought an unprecedented revolution in the past few decades, permeating every facet of our lives. From generating content to predicting market trends and redefining industries, AI seems to have no bounds.
But can it also help in screening the deadliest of diseases, like breast cancer?
Technology made its way into the healthcare long ago. Digital pathology has already been playing a pivotal role in modern medical practice for quite some time now. Using computerized technology, high-resolution digital images are being generated from the radiological films and glass slides of tissue specimens. This has not only made the data handling and storage easier but also allowed the pathologists and clinicians to better interpret the findings.
However, recent advancements in machine learning have now enabled the integration of artificial intelligence (AI) with digital pathology providing possibilities for image-based diagnosis.
Oncology is one of the medical fields which have benefited a lot from this advancement. And now, claims have been made that artificial intelligence could be the future of screening and risk stratification of breast cancer.
Let’s find out the truth behind that.
Breast Cancer Screening
Breast cancer is the most common cancer in women around the globe. According to the latest statistics provided by the World Health Organization (WHO), the disease took the lives of 685,000 women in the year 2020 while over 2.3 million new cases were diagnosed globally making it the world’s most prevalent cancer. In the United States, women have an overall 13% risk of developing breast cancer throughout their lifetime.
Various therapeutic modalities have been devised to treat breast cancer. However, the best way to tackle the disease is to catch it in the early stages. Breast cancer screening has, therefore, become a necessity rather than a choice in the lives of women.
Mammography, which is a low-dose X-ray imaging
of the breast, is currently the gold standard test for the screening of breast
cancer.
Experts recommend every woman should start having an annual mammogram after the age of 40. However, according to the American Cancer Society’s guidelines, women at high risk of developing breast cancer should undergo screening tests every year as soon as they turn 30.
(Check out the Cancer Screening Tests that should be on your Annual Health Checklist)
Integrating AI with Screening and Risk Prediction of Breast Cancer – An Efficacy Check
Since its huge success in various fields of life, artificial intelligence (AI) has been the pivot of focus of medical researchers, particularly in the fields of radiological imaging and diagnostics. The integration of AI in breast screening is one such example. However, AI has shown some amazing results in this new role. Claims have been made that using AI algorithms has a lot of potential and accuracy in detecting as well as predicting breast cancer in women.
Recently, a randomized-controlled study has been published in the Lancet Oncology Journal, comparing the efficacy and accuracy of AI-supported screening of breast cancer with that of the standard practice.1
This first-of-its-kind study was done on over 80,000 women living in Sweden. Two radiologists assessed half of the mammographic scans while the rest were analyzed by an AI system followed by the radiologist’s interpretation.
AI was able to detect breast cancer in 28% of the women – identifying 41 more cancer cases– as compared to the 25% found by manual screening by radiologists. Further, it also reduced the screen reading workload of the radiologist by 44%.
The high accuracy of cancer detection and reduced workload compelled the researchers to conclude that using AI in breast cancer screening is as good as two radiologists.
Similarly, in another study, researchers used the mammography AI algorithms for 5-years risk prediction of breast cancer and compared the risk scores with that of a conventional clinical model, i.e. BCSC (Breast Cancer Surveillance Consortium), which is based on accounting for the age, race, family history, birth history and density of the breast.
The study was published in the Radiology journal in June 2023 and reported that AI was better able to identify women at risk of developing breast cancer in the future. In comparison to the 21% of cancers detected by the clinical model, artificial intelligence predicted 28% of cancers by mere interpretation of the mammographic features of the breasts.
These findings imply that AI is capable of identifying patterns that are either imperceptible to the eye or require a great degree of training and expertise. Researchers concluded that integrating AI with clinical models could bring a revolution in the screening and risk stratification of breast cancers. 2
Yet, further large-scale studies are needed to confirm the efficacy of AI in this regard.
Challenges of Using AI in Breast Cancer Screening
Artificial intelligence (AI) offers numerous benefits, like better precision, faster results, lesser chances of error, and also decreased workload for diagnosticians, clinicians and pathologists. However, certain challenges are hindering the complete reliance on AI algorithms for breast cancer screening or any other diagnostic purposes.
Over-diagnosis
One big concern is the overdiagnosis of cancer. Experts say that where AI can identify the hard-to-catch malignancies, it also detects the relatively harmless lesions and labels them breast cancer which could lead to overdiagnosis and subsequent overtreatment. This can not only cause unnecessary tension to the patient but also increase the financial burden on the healthcare system.
Data Disparity
The breast cancer risks vary among people belonging to different races, ethnicities, and socioeconomic conditions. Therefore, it is currently impossible to apply a common ‘code’ to everyone unless data of all populations is universally integrated into the system.
Image-Based Diagnosis
The current AI models are centered on image data for diagnosis. This puts the diagnostic capabilities of the AI very much dependent on the quality of images.
Besides, the underutilization of patient histories, examination findings as well as health records further puts a check on the screening and diagnostic power of AI.
Ethical Consideration
AI also poses a lot of ethical risks as well, including loss of data confidentiality, data breach, privacy violation, and autonomy and consent issues of the patients.
What is the Future of Artificial Intelligence in Breast Cancer Screening and Prevention?
Artificial intelligence is already being used in women’s healthcare and has a lot of future potential. However, it will first assist the clinicians rather than replace them.
“Currently, there have been some advancements in various forms of computer-aided detection, but no one is relying on these systems alone”, say experts. Human intelligence is ultimately required for the final consideration and interpretation of results.
According to a report published in 2019 in the Future Healthcare Journal, 3
“For widespread adoption to take place, AI systems must be approved by regulators, integrated with EHR systems, standardized to a sufficient degree that similar products work similarly, taught to clinicians, paid for by public or private payer organizations and updated over time in the field.”
Bottom Line
Artificial Intelligence (AI) has shown a high efficiency in screening and risk stratification of breast cancer in women. However, addressing challenges, like overdiagnosis, data disparity, limited access to information, and ethical issues, is crucial for its successful integration. The collaborative synergy between AI and human expertise will shape the future of breast cancer care, ensuring accurate diagnosis and ethical healthcare practices.
References
- Lång K, Josefsson V, Larsson AM, Larsson S, Högberg C, Sartor H, Hofvind S, Andersson I, Rosso A. Artificial intelligence-supported screen reading versus standard double reading in the Mammography Screening with Artificial Intelligence trial (MASAI): a clinical safety analysis of a randomised, controlled, non-inferiority, single-blinded, screening accuracy study. The Lancet Oncology. 2023 Aug 1;24(8):936-44.
- Arasu VA, Habel LA, Achacoso NS, Buist DS, Cord JB, Esserman LJ, Hylton NM, Glymour MM, Kornak J, Kushi LH, Lewis DA. Comparison of Mammography AI Algorithms with a Clinical Risk Model for 5-year Breast Cancer Risk Prediction: An Observational Study. Radiology. 2023 Jun 6;307(5):e222733.
- Davenport T, Kalakota R. The potential for artificial intelligence in healthcare. Future healthcare journal. 2019 Jun;6(2):94.
Hello fitness enthusiasts! I am Dr. Haris, a dedicated certified medical practitioner, and a professional freelance medical writer with many years of experience in medical writing. My journey includes a noteworthy stint as a former house surgeon at Mayo Hospital Lahore, and now I am pursuing residency training in the field of cardiac surgery.
Being a doctor, I always find immense joy in sharing valuable insights, guiding individuals on the path to optimal health. I am proud to be a part of the Blufashion team, where I contribute by reviewing and writing articles related to everyday health and wellness. Our commitment is to empower Blufashion readers by disseminating accurate health information.
Connect with Dr. Haris on LinkedIn.
