Forward-thinking authors Drs. Saurabh Jha and Eric Topol present an interesting view of how advances in artificial intelligence (AI) could impact the roles of pathologists and radiologists.(1) Because pathology and radiology have many tasks that can be handled by AI, Drs. Jha and Topol posit that in the future, these two specialties might merge into a new role, the “information specialist,” whose job will be managing the information that AI extracts from images and associating it with the patient’s clinical picture.
Computers That Can Learn Can Perform Some Tasks Better Than Humans
AI refers to computers that have the ability to “think” and learn, and is gaining traction in medicine. Drs. Jha and Topol believe that because of the continued growth in big data, AI has the power to take over certain tasks where computers outperform humans, particularly in image-rich fields such as pathology and radiology. For example, a trauma radiologist might review up to 4,000 CT images, which can lead to eye fatigue; AI software can scan images endlessly and not lose any acuity or become fatigued. Similarly, pathologists can suffer from fatigue and eye strain as they scan images to detect medical anomalies.
Natural Language Processing & Machine Learning Guide Cancer Treatments
AI is already capable of scanning radiographs to identify fractures, and has developed its own internal algorithms to identify fractures using deep learning.(1) IBM Watson is another example of AI; it can identify pulmonary embolism on a CT scan and detect abnormal wall motion on echocardiography. Additionally, Watson’s natural language processing and machine learning is being used at Memorial Sloan Kettering Cancer Center to aid in assessment of large amounts of structured and unstructured medical record data to guide treatment recommendations for breast, lung, and colorectal cancers.(2)
Artificial Intelligence Redefining Roles in Medicine
As new technologies become available, roles become redefined and humans can be better occupied in tasks that require human traits. Furthermore, as imaging data accumulates beyond the amount that can be processed by the human eye, it can be easily supported with AI abilities. AI can be used to support the move toward personalized medicine; it should not be viewed as job replacement, but as a complementary tool for providers to use in those specific areas where AI can outperform humans.
1. Jha, S., & Topol, E. (2016). Adapting to artificial intelligence: Radiologists and pathologists as information specialists. Journal of the American Medical Association.
2. Monegain, B. (2016, December 9). IBM Watson accurately matches oncologists’ advice, study finds. Retrieved from http://www.healthcareitnews.com/news/ibm-watson-accurately-matches-oncologists-advice-study-finds?
Orchard is committed to keeping you informed and being a trusted resource that you can turn to for industry-related education. As always, we welcome your feedback. Follow us on Twitter at @orchardsoftware, and feel free to respond to this post by emailing us at news@orchardsoft.
Kim Futrell, BS, MT(ASCP)
Products Marketing Manager
Orchard Software Corporation