The Watson supercomputer, created by technology company IBM, uses the power of artificial intelligence (AI) in ways that benefit mankind. In 2011, the world was introduced to the potential of AI when Watson won the title of world champion on the popular game show Jeopardy! In 2011.
Since then, Watson’s developers have focused on customizing the AI for medical use. Programmers have integrated thousands of patient records, scores of medical texts, and all the content in the PubMed and Medline databases into Watson’s working memory. In all, the AI has consumed 2 million pages of text that represent over 600,000 pieces of medical evidence, 25,000 training cases, and 14,700 hours of clinical training.
The Benefits of Using AI in Medical Diagnoses
AI programs, like Watson, are gaining importance in information-heavy fields, like medicine. The supercomputer is able to store more information than a human brain and can access the information more quickly. In addition, a computer-generated diagnosis is based on evidence and established fact, while human doctors may be swayed by cognitive biases or confidence issues.
The Watson AI uses an interface that allows users to speak to it in a natural way. The machine is able to gather and analyze information from patients and compare it to stored information. Based on this inquiry, the AI can then make an accurate diagnosis.
More than just comparing bits of information, the Watson supercomputer is able to form hypotheses independently. In a process that is similar to the way humans learn, the machine can then test and evaluate its own hypotheses, disregarding what doesn’t work and adjusting for error.
In 2011, McAfee wrote about the advantages of using AI for medical purposes. AI diagnoses are based on the complete published body of medical knowledge, making it the most up-to-date medical resource. Diagnostic programs are also accurate, consistent, cost very little to operate, and can be used anywhere in the world.
AI: The Future in Medical Care
When using AI for diagnostics, a physician describes symptoms and other contributing factors to the program. The computer then mines patient history and its own databases, and along with test results and other pieces of medically relevant data, creates a hypothesis. The hypothesis is tested against its information on treatments, hospital records, and published studies and research. Once satisfied that the hypothesis is valid, Watson can then give a customized treatment plan for the patient.
While Watson has incredible capacity, only a few companies are taking advantage of it. MD Anderson Cancer Center uses AI to develop treatment plans for leukemia patients. The medical company WellPoint uses Watson-based software to allow patients to evaluate their doctor’s treatment plans.
Currently, the Watson AI is not able to fully use all of the information it has access to. Developers are working with the program to help it develop the mental agility and versatility of thought that makes human doctors so effective.
“It is good to know that a doctor of osteopathic medicine or DO physician is supervising IBM Watson. This is because DOs have a more balanced perspective of health in that they are trained in conventional medicine as well as a form of alternative medicine known as “osteopathic manipulation.” In other words, osteopathic physicians have as part of their basic medical philosophy that the human body has the innate ability to heal itself when somatic dysfunctions are corrected. This is important for Watson because such a physician is less likely to be biased in his or her treatment approaches of the patient and thus, will not limit options available to Watson for clinical recommendations. Therefore, we the people benefit from Watson not being limited to providing us with only drugs and surgeries as treatment options as would be the case if an allopathic physician (a.k.a., Medical Doctor or “MD”) was supervising IBM Watson.” ~ Daniel F. Royal, DO, HMD, JD Turtle Healing Band Clinic, Las Vegas, NV
