It seems straight out of a nightmare: artificial intelligence (AI) ruling the world. Several movies have been based on the negatives of AI: AI turning against humans or AI leading to the dehumanization of the world. However, this portrayal of AI is misleading. AI is not disruptive to human life; in fact, it may be the next biggest advancement. AI is on the brink of revolutionizing medicine, as it has a very broad range of potential applications in healthcare.
What is Artificial Intelligence?
Artificial intelligence is the theory and development of computer-based systems that can execute tasks generally requiring human intelligence. The term “artificial intelligence” was first coined by John McCarthy, an American mathematician, in 1956 . However, AI research started after WWII and was first conceptualized by mathematician Alan Turing. The Turing test is used often in the field of AI; it is a test that determines if a machine is truly intelligent or not based on if the machine can fool an interrogator with humanoid behavior . This is helpful in many different fields and in many different capacities, such as healthcare.
Artificial Intelligence in Healthcare
One major aspect of healthcare in which AI will play a major role in coming years is data management. Each patient generates a lot of data, ranging from histories and vitals to procedure details and care plans. All of this metadata is hard to process, analyze, and transform into useful information . Also, if a patient is switching doctors or changing healthcare systems, it may take a long time to transfer the patient’s information to the new location, and even then, information might be missing. AI programs can speed up data transferring processes, help insurers detect fraud, and help physicians identify best practices. For example, Arkansas Data Network analyzes readmission and resource usage to develop better treatment protocol . Some of these programs, as part of the movement towards precision medicine, are also starting to incorporate whole genome sequencing to find new potential indicators of disease and improve treatment. In addition, the Google Deepmind Health program is working on making the transfer of data from doctor to doctor or hospital system to hospital system significantly more organized and uniform, which could make a huge difference in time-sensitive cases such as cancer treatments . However, issues with these data management technologies are that they are vulnerable to cyberattacks, and there is also a lengthy adoption process since data storage systems are quite disparate across healthcare systems.
Another potential application of AI in healthcare is in assisting with repetitive analytical tasks. This application is especially important to the field of radiology. Radiologists analyze various scans, such as X-rays, CT scans, and MRI scans, and then compare the scan to what the scan is supposed to look like, looking for concerning abnormalities and problems. When these abnormalities were taught to an AI program, the program was able to point them out on scans. One example of this technology is called Medical Sieve. It quickly points out disease-depicting regions and is capable of creating a concise summary of the anomaly from various viewpoints . Nevertheless, some worry that it is risky to solely trust technology, even if highly accurate, when it comes to human life. Thus, at least for now, radiologists will still play a major role in analyzing scans.
AI will also have major influence in the field of surgery. According to a Johns Hopkins study, surgical error (along with other medical errors) leads to over 250,000 deaths a year and is the 3rd leading cause of death in the U.S. . However, sometimes higher surgical precision and range than can be provided with the human hand can prevent such errors. The da Vinci Surgery is a minimally invasive system of small, wristed instruments that the surgeon can control. These instruments can operate through small incisions, have magnified 3D high-definition vision, and can bend and rotate further than the human hand . These tools are not meant to replace the role of a surgeon but instead, serve as another tool that surgeons can use if needed, especially in surgeries that require higher mobility or a greater field of vision. Nevertheless, in the future, tools that can work more independently may be developed. One major obstacle impeding the growth of surgical technology is that artificial intelligence requires practice and past experience — information fed into the program — to function accurately. For example, in order to train the Android phone on voice recognition, it was trained on 10,000 hours of annotated speech . Unfortunately, in medical AI, the training process can be risky to human lives, especially with programs that are directly connected to patient care, such as those utilized in surgery.
Artificial Intelligence in the Future
The application of artificial intelligence in medicine has the potential to benefit patients, physicians, and healthcare administrators. This technology is in the process of enabling healthcare delivery to become faster, more efficient, and less error-prone.
However, many people have reservations regarding these changes. A major concern is the dehumanization of medicine since the patient-doctor relationship is a key aspect of patient care. Would the further advancement of AI slowly limit the role of physicians in healthcare? Will online consultation-based AI take over the role of primary care physicians, analyzing software like Medical Sieve take over the role of radiologists, and robotic hands take over the role of surgeons? If this were to happen, would patient care be improved due to lower medical error and faster care?
Many of these reservations stem from apprehension of technological advancements. In reality, as AI and other forms of technology advance, physicians will learn how to integrate it into their existing practices. By complementing and improving different aspects of medicine, AI will revolutionize the delivery of healthcare.
- “The True Father of Artificial Intelligence.” OpenMind, OpenMind, 31 Aug. 2016, www.bbvaopenmind.com/en/the-true-father-of-artificial-intelligence/.
- McCarthy, John. “WHAT IS ARTIFICIAL INTELLIGENCE?” Stanford Computer Science Department, 12 Nov. 2007, www-formal.stanford.edu/jmc/whatisai/. Accessed 29 Sept. 2017.
- Koh, Hian Chye, and Gerald Tan. “Data Mining Applications in Healthcare.” Journal of Healthcare Information Management, vol. 19, pp. 64–72.
- “Welcome to DeepMind Health.” DeepMind, Deepmind Technologies Limited, 2017, deepmind.com/applied/deepmind-health/. Accessed 29 Sept. 2017.
- “Medical Sieve.” Medical Sieve – IBM, IBM, 25 July 2016, researcher.watson.ibm.com/researcher/view_group.php?id=4384. Accessed 29 Sept. 2017.
- “Study Suggests Medical Errors Now Third Leading Cause of Death in the U.S.” Johns Hopkins Medicine, Johns Hopkins Medicine, 3 May 2016, www.hopkinsmedicine.org/news/media/releases/study_suggests_medical_errors_now_third_leading_cause_of_death_in_the_us. Accessed 29 Sept. 2017.
- “Da Vinci® Surgery: Minimally Invasive Surgery.” Da Vinci Surgery | Robotic-Assisted Surgery, Nov. 2015, www.davincisurgery.com/. Accessed 29 Sept. 2017.
8. Heigold, G., et al. “Multilingual acoustic models using distributed deep neural networks.” 2013 IEEE International Conference on Acoustics, Speech and Signal Processing, 2013, doi:10.1109/icassp.2013.6639348.