William Sander, Class of 2022
Imagine going to the doctor’s office for a simple test and walking out an hour later knowing what possible diseases you might contract over the course of your lifetime, and possibly how long you might live. This seems like a frightening and futuristic scenario that you might expect to be on Netflix’s Black Mirror, but the reality is that this kind of technology is already being created. Artificial Intelligence (AI) has been receiving a lot of coverage lately from some claiming that it is the end of society as we know it and from others who claim that it is something to be embraced and even taught, as is the case with the Massachusetts Institute of Technology’s plans to spend one billion dollars to erect a college for AI education. AI is an abstract notion that is troubling and vexing to the vast majority of people, but what cannot be denied is its inherent usefulness to the field of medicine. Many researchers at the forefront of the Artificial Intelligence era believe that we are beginning to experience the “Fourth Industrial revolution” (Medical Futurist, 2018). What is different about this new revolution though? Klaus Schwab, the author of the The Fourth Industrial Revolution, claims that while the other industrial revolutions have been characterized by animal power, industrial production, and large technological production, this fourth industrial revolution is “fusing the physical, digital and biological worlds, impacting all disciplines, economies, and industries” (Schwab, 2016). The obvious benefits of increased efficiency and more widespread sharing of information that come with Artificial intelligence are shadowed by the ominous notion that we, as humans, might become replaced, and have our professional jobs rendered useless by the unquestionable potential that artificial intelligence has on many fields. One such field in which the potential of Artificial Intelligence seems to be the greatest is in Radiology.
In addition to performing tests such as CT scans, X-Rays, Ultrasounds, and MRIs, Radiologists are also responsible for interpreting the results of such tests and formulating a diagnosis of medical conditions exhibited by the patient. Artificial Intelligence tools would be able to record and interpret patterns in data from millions of sample scans, all from different patients, and be able to predict outcomes and diagnoses of individual tests. This function alone is an immensely-powerful tool for propelling the field of Radiology far into the future of medicine and medical technologies. In fact, researchers at the University of Adelaide are using Artificial Intelligence to test how effectively it can predict death within the next five years, and with only sample CT scans from forty-eight patients, an AI-programmed computer predicted outcomes with about seventy percent accuracy, and it nearly matches the accuracy with which radiologists were able to make a correct diagnosis (Medical Futurist, 2018). Furthermore, these so-called, “deep learning systems”, are programmed to interpret up to sixteen-thousand data points, collected from a myriad of scans, about the condition of a patient. These interpreted scans are expected to be used in the near future to give medical practitioners better insight into the condition of their patient. These are just two examples of the thousands of applications Artificial Intelligence has to the field of radiology.
This new reality of increased proficiency with information processing in medicine has not been embraced by Radiologists. In fact, Stanford Professor of Bioinformatics, Researcher, and Radiologist, Curtis Langlotz, discussed how many students, who are pursuing a career in the field of Radiology, are concerned about whether Radiology is still even a viable profession. These students are fearing that the advances in AI technology in the next ten years might be enough to render them jobless (Medical Futurist, 2018). These concerns over being replaced by AI are largely invalid. Despite what most people might assume, AI technology does not aim to replace, but merely to be used as a tool by professionals to help them treat patients. AI technology can only be used to identify certain objects or shapes on a scan, similar to how spell check works and can make a prediction of a condition, but it is a prediction that still must be validated by a professional. Even further, processes for gathering this information and determining how to interpret it must also be evaluated by radiologists every step of the way as radiologists do not simply identify spots of possible risk in scans, but they “synthesize the objects in a scan into a coherent, meaningful interpretation: the patient’s story” (Shaffer, 2018). The point to take away is that the job of radiologists is no less essential. In fact, it is now more essential, as the field of Radiology begins to fuse with AI technologies. A computer might be able to recognize certain shapes in scans, but this process pales in comparison to being able to identify a problem, with the level of complexity that an experienced Radiologist has. What many tend to forget when considering the future of Artificial Intelligence and the roles it will play in medicine is the fact that fields like Radiology have always been constantly evolving and adapting new methods of medical analysis and that Radiology now is absolutely nothing like what it was ten years ago. AI technology must be accepted, not pushed away, in the advancing fields of Medicine, and especially in the field of Radiology.
References
In addition to performing tests such as CT scans, X-Rays, Ultrasounds, and MRIs, Radiologists are also responsible for interpreting the results of such tests and formulating a diagnosis of medical conditions exhibited by the patient. Artificial Intelligence tools would be able to record and interpret patterns in data from millions of sample scans, all from different patients, and be able to predict outcomes and diagnoses of individual tests. This function alone is an immensely-powerful tool for propelling the field of Radiology far into the future of medicine and medical technologies. In fact, researchers at the University of Adelaide are using Artificial Intelligence to test how effectively it can predict death within the next five years, and with only sample CT scans from forty-eight patients, an AI-programmed computer predicted outcomes with about seventy percent accuracy, and it nearly matches the accuracy with which radiologists were able to make a correct diagnosis (Medical Futurist, 2018). Furthermore, these so-called, “deep learning systems”, are programmed to interpret up to sixteen-thousand data points, collected from a myriad of scans, about the condition of a patient. These interpreted scans are expected to be used in the near future to give medical practitioners better insight into the condition of their patient. These are just two examples of the thousands of applications Artificial Intelligence has to the field of radiology.
This new reality of increased proficiency with information processing in medicine has not been embraced by Radiologists. In fact, Stanford Professor of Bioinformatics, Researcher, and Radiologist, Curtis Langlotz, discussed how many students, who are pursuing a career in the field of Radiology, are concerned about whether Radiology is still even a viable profession. These students are fearing that the advances in AI technology in the next ten years might be enough to render them jobless (Medical Futurist, 2018). These concerns over being replaced by AI are largely invalid. Despite what most people might assume, AI technology does not aim to replace, but merely to be used as a tool by professionals to help them treat patients. AI technology can only be used to identify certain objects or shapes on a scan, similar to how spell check works and can make a prediction of a condition, but it is a prediction that still must be validated by a professional. Even further, processes for gathering this information and determining how to interpret it must also be evaluated by radiologists every step of the way as radiologists do not simply identify spots of possible risk in scans, but they “synthesize the objects in a scan into a coherent, meaningful interpretation: the patient’s story” (Shaffer, 2018). The point to take away is that the job of radiologists is no less essential. In fact, it is now more essential, as the field of Radiology begins to fuse with AI technologies. A computer might be able to recognize certain shapes in scans, but this process pales in comparison to being able to identify a problem, with the level of complexity that an experienced Radiologist has. What many tend to forget when considering the future of Artificial Intelligence and the roles it will play in medicine is the fact that fields like Radiology have always been constantly evolving and adapting new methods of medical analysis and that Radiology now is absolutely nothing like what it was ten years ago. AI technology must be accepted, not pushed away, in the advancing fields of Medicine, and especially in the field of Radiology.
References
- Schwab, K. (2016, January 11). The Fourth Industrial Revolution. Retrieved November 10, 2018, from https://www.weforum.org/pages/the-fourth-industrial-revolution-by-klaus-schwab/
- Shaffer, P. (2018, July 17). Opinion: No, AI Will Not Replace Radiologists. Retrieved November 10, 2018, from https://www.the-scientist.com/news-opinion/opinion--no--ai-will-not-replace-radiologists-64506
- The Medical Futurist. (2018, July 26). Artificial Intelligence Will Redesign Healthcare. Retrieved November 10, 2018, from https://medicalfuturist.com/artificial-intelligence-will-redesign-healthcare
- The Medical Futurist. (2018, January 15). The Future of Radiology and Artificial Intelligence. Retrieved November 10, 2018, from https://medicalfuturist.com/the-future-of-radiology-and-ai
Proudly powered by Weebly