Artificial Intelligence in Medicine
From the development of MRI and X-ray machines to the applications of advanced prosthetics and laser surgeries, the progression of the healthcare industry has been revolutionary. However, we are on the dawn of a new era in healthcare as the complexity and increase of data has necessitated the use of artificial intelligence (AI). The potential of AI in medical aspects such as operation, administration, patient interaction, and research (to name a few) is immense, but the fact remains that healthcare is still a human-dominated industry. And there are legitimate arguments for why this industry should remain this way. Ultimately, the ethical concerns of AI implementation into the healthcare industry such as machine’s lack of accountability and potential displacement of current healthcare workers must be taken into account before AI is fully integrated into the healthcare industry. Therefore, AI should be assimilated into the health care industry on a case by case basis. This article will detail several healthcare applications of AI that will yield tremendous benefit to the industry with minimal drawbacks.
AI in Patient Engagement
Doctors can only prescribe health care plans for patients to carry out independently; it is solely up to the patients to follow through on the plans and take actions in the best interests of their health. Therefore, encouraging patients to take preemptive health measures for themselves is a current challenge in the medical field. AI should be utilized in this aspect because of its ability to harness machine learning, specifically supervised learning. This allows AI to input data on patient behavior, characteristics, and situation in order to predict what treatment protocols are most compatible for each patient. With health care plans tailored to individual patient circumstances rather than a generic script, patients will be more likely to follow through on them. For example, a patient with asthma is more likely to stick with yoga training than a rehab therapy involving cardio. Along with bolstering doctors’ abilities to prescribe custom treatments, AI is immensely valuable for individuals looking for a quick diagnosis. In fact, such technology is already available: Buoy Health is an AI-based symptom and cure checker that uses a chatbot that listens to patient symptoms and health concerns to generate a health care plan based on its diagnosis.¹ Buoy’s symptom checker was built to empower its users with better understanding their medical symptoms. No more endless Googling. No more guessing. No more doctor visits just to be told you have a stomach flu. Since such technology is already available, the next steps are refining its credibility and making it more accessible. With AI that helps patients curb minor health concerns before they even arise, individuals have more control over their own health, and patients receive immediate cures rather than being forced to wait in emergency rooms.
AI in Surgery
Human surgical performance is dictated by multiple physical, mental, and technical factors, meaning that there is certainly a degree of variability in the execution of each surgery. In surgery situations where lives hang in the balance, it is essential that variability is minimized. For this reason, it is logical to introduce machines into the equation, particularly ones enhanced by AI. AI operating machines have several advantages over humans including fatigue resistance, less technical errors, and wider ranges of motion in addition to having the human ability to learn over time. It is important to note that since most of these devices rely on manual operation, the human aspect of conducting surgeries will not be removed. Ultimately, human surgeons still retain the ability to make the decisions on surgery procedure, and their capabilities are greatly enhanced thanks to AI abilities. Limited use of AI surgical robots has already allowed for new medical revolutions such as minimally invasive surgery. With this methodology, doctors do not have to operate on patients through large incisions; instead the robots operate using only minimal incisions. AI working tandem with human surgeons, not only maintains the long-term value of human surgeons, but gives them the ability to conduct a wider range of error-free surgeries.
AI in Research
There are countless discoveries waiting to be made in the medical field, and integrating AI into research is the key to unlocking many of them. AI’s most useful ability in this context is unsupervised learning; in other words, AI can sift through large cascades of unsorted data and identify patterns that have escaped the human eye. One company currently utilizing AI for this purpose is Tempus, which is using AI to analyze the world’s largest collection of clinical and molecular data in order to personalize healthcare treatments, uncover new treatments, and find disease cures.² Often times, human medical discoveries occur by accident, but with the use of AI in the research, discovery can be reframed into a process that quantifies of previously unnoticed patterns. Another reason AI will be so useful in the research industry is its image recognition ability. Current AI algorithms are already outperforming radiologists at spotting malignant tumors, for example, according to the Future Healthcare Journal.³ Other than time and money, AI used in research has no downsides as theoretical ideas can be tested in isolated environments. Furthermore, the payoff of AI in research is potentially new medical advancements, justifying any potential pitfalls. Once research actually stops being research and moves into implementation stages, ethical considerations can be reevaluated.
AI in Administration
In all likelihood, filling out paperwork is not the highlight of anyone’s job, and this goes double for many workers in the medical field. In fact, the average US nurse spends 25% of work time on regulatory and administrative activities according to a study conducted by Ann Hendrich and published in the Permanente Journal.⁴ Time spent completing repetitive manual labor takes a large chunk of nurses’ time and prevents them from gaining more medical experience. With the help of AI, this busy work can be handled automatically, more efficiently, and more cheaply allowing nurses to have more time to prioritize patient service and shoulder higher-level responsibilities. Reaching this goal is the platform of the AI tool Olive, which was designed to perform the healthcare industries most repetitive tasks and cut costs.⁵ The emphasis on reducing costs is especially essential because according to the New England Journal of Medicine, currently 30% of healthcare costs are associated with administrative tasks.⁶ Since AI like Olive can handle everything from eligibility checks to unadjudicated claims and data migrations at a higher efficiency than humans, not only are human resources being wasted in this line of work, but they are actually comparatively inefficient. The automatization of paperwork and other minor administrative tasks will also improve the healthcare official on standby to patient ratio, and with the unpredictable nature of health scares and accidents, the more nurses available the better.
Conclusion
A balance must be maintained between AI’s potential and the ethical dilemmas that arise as a result of its implementation. Partial integration is the most practical way to implement AI into the healthcare industry because medical workers are not rendered expendable and their skills are enhanced with the use of AI. The implications of this shift would be massive as this would change the way medical professionals are trained before they enter the medical field, but there would be just as many questions if we did not use the technology available to us. After carefully weighing the potential and the negative effects of AI in the medical industry against each other, we must integrate AI into healthcare aspects where it can be the most useful.
[1]: “The Top 12 Health Chatbots.” The Medical Futurist, The Medical Futurist, 20 Jan. 2020, medicalfuturist.com/top-12-health-chatbots/.
[2]: Daley, Sam. “Surgical Robots, New Medicines and Better Care: 32 Examples of AI in Healthcare.” Built In, Built In, 4 July 2019, builtin.com/artificial-intelligence/artificial-intelligence-healthcare.
[3]: Davenport, Thomas, and Ravi Kalakota. “The Potential for Artificial Intelligence in Healthcare.” Future Healthcare Journal, vol. 6, no. 2, 2019, pp. 94–98., doi:10.7861/futurehosp.6–2–94.
[4]: Hendrich, Ann. “A 36-Hospital Time and Motion Study: How Do Medical-Surgical Nurses Spend Their Time?” The Permanente Journal, 2008, pp. 25–34., doi:10.7812/tpp/08–021.
[5]: Landi, Heather. “Healthcare Has a Lot of Tedious, Repetitive Tasks. This CEO Is Using AI to Fix That.” FierceHealthcare, FierceHealthcare, 26 Sept. 2019, www.fiercehealthcare.com/tech/yale-new-haven-centura-health-working-ai-company-olive-to-tackle-administrative-burdens.
[6]: Woolhandler, Steffie, et al. “Costs of Health Care Administration in the United States and Canada.” New England Journal of Medicine, vol. 349, no. 8, 2003, pp. 768–775., doi:10.1056/nejmsa022033.