Should you get a second opinion from Artificial Intelligence (AI)? With the rampant advancement of AI technology and robotics, even healthcare professionals are concerned about the possibility of losing their jobs.
It all started at Bács-Kiskun County Hospital in Budapest, Central Hungary. Dr. Éva Ambrózay, a radiologist with over 20 years of experience, was reviewing the details of a patient’s mammogram results. According to two other radiologists who examined the file, the patient was in the clear, showing no signs of tumor growth. However, as AI software detected abnormalities that the two human radiologists did not, it prompted Dr. Ambrózay to re-evaluate the mammogram findings. Low and behold, the machine was onto something. The patient may have breast cancer.
Can AI Detect Cancer That Doctors Miss?
AI advancements and machine learning have progressed by leaps and bounds in recent years, especially in 2023. Artificial intelligence is now capable of detecting tumors that medical professionals overlooked. These deep machine-learning systems can identify cancer signs proficiently, and sometimes more effectively than human radiologists. Nonetheless, cancer detection systems still have hurdles to overcome before widespread implementation.
Early Beginnings of AI Software for Cancer Detection
Central Hungary is notable for its advanced breast cancer screening programs. The country has one of the world’s largest testing grounds for AI cancer detection technology.
Hungarian doctors began testing and using AI in the medical industry in 2021. Today, those AI systems aid in pinpointing tumors that a human radiologist failed to detect. Health facilities in the United States, European Union, and Britain support Hungary’s efforts by initiating tests and providing clinical data to help improve cancer detection technology.
Case Study on AI-Supported Cancer Screening
In August 2023, the results of the first randomized controlled trial regarding AI cancer detection systems became available in THE LANCET Oncology. The year-long study, which involved 80,033 Swedish women aged 54, compared AI-supported cancer screening with conventional methods.
Two human radiologists assessed half of the scans, while AI-supported screening analyzed the remaining half. After which, one to two radiologists interpreted the AI-assessed data.
The Results
Overall, 244 participants (28%) recalled from AI-supported cancer screening showed signs of cancer compared with 203 participants (25%) recalled from conventional screening. Support from AI technology found 41 more cancer cases. Of which, 22 have not spread or metastasized (in situ), and 19 have invaded other tissues or organs (invasive).
Radiologists in the AI group completed 36,886 fewer screen readings than the conventional care group. AI assistance reduced the radiologists’ screen-reading workload by 44%. As for the false-positive rate, it was 1.5% in both groups.
The researchers concluded, “AI-supported mammography screening resulted in a similar cancer detection rate compared with standard double reading, with a substantially lower screen-reading workload.” In other words, AI can help radiologists speed up the cancer detection process.
Other Developments and Ongoing Studies
As researchers uncover more breakthroughs, AI may improve public health and pioneer a new era in cancer prevention and detection.
- AI-assisted prostate cancer detection: Research shows that AI technology can accurately classify cancer grades, such as the Gleason scoring system for prostate cancer and lymph node metastasis identification.
- AI-assisted lung cancer detection: Another area where artificial intelligence and deep learning technology can aid cancer research is integrating extensive data analysis with pathology assessment and/or image review. For instance, in a 2022 study published in the JAMA Network, the researchers used AI technology to analyze patients’ medical records and chest X-ray results. In conclusion, AI helped identify individuals at high risk for lung cancer.
- AI-assisted gene mutation presage: Science and tech professionals are also exploring AI’s potential in predicting cancer-causing gene mutations from histologic analyses.
Related: Artificial Intelligence and Cancer (Part 3): The Current State of Cancer Risk Prediction
AI-Assisted Breast Cancer Detection Goals for the Future
Despite its existing achievements, cancer detection technology has a long way to go. Additional clinical trials are necessary before AI systems become widely accepted as automated breast cancer screening tools beyond Hungary and other territories using AI technology. With breast cancer, the primary goal is for AI-powered cancer detection systems to generate accurate and reliable results for female patients of all ages, ethnic backgrounds, and body types. The system must also recognize aggressive and rare breast cancer types, as well as minimize or eliminate false-positives readings.
Machine vs. People: Concerns From Doctors and Patients
Similar to professionals from other industries, AI advancement has sparked discussions about the odds of replacing human healthcare providers. Countless programmers, graphic designers, writers, artists, and other experts are already fighting against the AI takeover. At the same time, AI developers are facing backlash and resistance from some medical providers and healthcare organizations.
Similar to other professions, those fears and uncertainties have slightly calmed. Many experts believe that regardless of how far AI technology progresses, patients will mostly trust AI-generated results when in partnership with human practitioners.
At the end of the day, AI cancer detectors could save millions of lives, said Dr. László Tabár, a professor of radiology at Uppsala University in Sweden. He praised AI and deep learning, saying the technology impressed him after reviewing its capabilities in breast cancer screening.
The Bottom Line
No matter how groundbreaking, these discoveries have yet to scratch the surface of advancing cancer research via artificial intelligence. Although the potential applications seem endless, as of August 2023, we have only made progress in the field of cancer imaging.
The road ahead is long, complex, and unpredictable. Still, as researchers press on and continue developing tools that make screening faster and more accurate, there is hope for a significantly brighter future for all – one where medical professionals and AI technology work hand in hand in the battle against cancer.
About New Hope Medical Center
Count on us to be by your side as you fight and survive cancer.
The New Hope Unlimited team stays abreast of cutting-edge cancer research and treatment innovations. Our pledge to patients includes access to some of the most up-to-date, powerful, and comprehensive options available. If you seek guidance or a consultation for your illness, our doors are open. Reach out to us today at 480-757-6573. You may also complete this online form to contact our operators.
Sources
[1] Satariano, A., Metz, C., & Times, A. S. F. N. Y. (2023, March 6). How A.I. is being used to detect cancer that doctors miss. The New York Times. https://www.nytimes.com/2023/03/05/technology/artificial-intelligence-breast-cancer-detection.html
[2] Carter SM, Rogers W, Win KT, Frazer H, Richards B, Houssami N. The ethical, legal and social implications of using artificial intelligence systems in breast cancer care. 2020 Feb;49:25-32. doi: 10.1016/j.breast.2019.10.001. Epub 2019 Oct 11. PMID: 31677530; PMCID: PMC7375671.
[3] Busby D, Grauer R, Pandav K, et al. Applications of artificial intelligence in prostate cancer histopathology. Urol Oncol. 2023;11:S1078-1439(22)00487-2. doi: 10.1016/j.urolonc.2022.12.002
[4] Sandbank, J, Bataillon, G, Nudelman, A, et al. Validation and real-world clinical application of an artificial intelligence algorithm for breast cancer detection in biopsies. npj Breast Cancer. 2022;8(129). doi: https://doi.org/10.1038/s41523-022-00496-w
[5] Alam MR, Seo KJ, Abdul-Ghafar J, et al. Recent application of artificial intelligence on histopathologic image-based prediction of gene mutation in solid cancers. Brief Bioinform. 2023 Apr 27:bbad151. doi: 10.1093/bib/bbad151
[6] Raghu VK, Walia AS, Zinzuwadia AN, et al. Validation of a deep learning-based model to predict lung cancer risk using chest radiographs and electronic medical record data. JAMA Network. 2022;1;5(12):e2248793.
