By guest author: Brian Kaplan, M.D. – Chair, Department of Otolaryngology-Head and Neck Surgery at Greater Baltimore Medical Center Director, Cochlear Implant Program at Greater Baltimore Medical Center
Artificial Intelligence (AI) has been around for some time. Every day we see another story about the use of artificial intelligence, machine learning or big data in the news. Advances in medicine have often relied on new discoveries in science and now healthcare is being transformed by the use of these advanced technologies.
In 1959, Arthur Samuel was credited with being the first person to realize that we could not teach a computer everything we know in order to complete given tasks. We needed to teach computers how to learn for themselves1. This discovery, along with the advent of the internet, set into motion the ability for computers to take the next step into current state-of-the-art AI and machine learning. In the 1960s, researchers developed the first problem-solving systems in organic chemistry, which is considered the first application of artificial intelligence. By the 1970s, researchers moved from basic science to applications in medicine with attempts to automate diagnosis2. Today, most AI employed in healthcare revolves around classification using deep learning technologies. It is not that AI systems cannot diagnose certain conditions, but the requirement for machine learning has changed from high cognitive function to classifications tasks. This step in the use of technology requires convincing humans that the classifications are accurate, consistent and repeatable. In a recent study, AI systems were found to be able to identify melanomas (through classification) at a higher rate than expert dermatologists and with greater sensitivity and specificity3. However, the most successful AI solutions are those that augment human action rather than replace it.
There was a time when some experts predicted the use of robots in all areas of healthcare, specifically in surgery. Granted there are some rote aspects of surgery that a robot can certainly handle, but the state-of-the-art techniques employed today are robot-assisted surgery. In LASIK surgery, a skilled surgeon uses a robotic device to complete most of the procedure. The surgeon is still responsible for all of the measurements and calculations required during the surgical procedure and is certainly the crucial asset in the case of an emergency or unexpected anatomy. Machines have an innate advantage during rote tasks of being fast, accurate and operate with a lack of bias, but human discernment provides a successful outcome when unanticipated events occur. Robotic capabilities in surgery may eventually allow remote-robotic surgery. This could change clinical care by providing care to patients who may not otherwise receive advanced care. This increase in access to care is beneficial to remote patients and the clinicians who care for them.
New AI tools help medical professionals change their clinical practice and to standardize care. Allowing AI systems to assist professionals can not only standardize care, it can increase productivity. The key to successful use of AI systems is to ensure the human factor never fades from the healthcare appointment. In the study mentioned previously, once the dermatologists were provided with additional information from the patients, their diagnostic accuracy increased and the article concludes that there was no substitute for a thorough clinical examination to provide a full healthcare picture for patients and their providers.
AI is integrated into many aspects of our lives and we should expect that to continue. Growth in the healthcare sector continues to be strong with one source estimating that artificial intelligence in US healthcare will continue at 35% rate for the next eight years4. Many predictors bemoan the hurdles needed to fully integrate AI into mainstream healthcare. Using AI to standardize care provides assistance to professionals, which allows for enhancing the lives of patients and healthcare professionals.
See the references below for more information.
About our guest author: Brian Kaplan, M.D. – Chair, Department of Otolaryngology-Head and Neck Surgery at Greater Baltimore Medical Center Director, Cochlear Implant Program at Greater Baltimore Medical Center
Dr. Kaplan was born and raised in Boston, Massachusetts. He attended Cornell University before enrolling at the University of Virginia for both medical school and residency training. He currently serves as the Chairman of the Department of Otolaryngology, and Director of the Cochlear Implant Program, at the Greater Baltimore Medical Center.
Past research interests have included cartilage and inner ear hair cell regeneration. Clinical studies have involved various ototopical agents, along with hybrid and traditional cochlear implantation.
Dr. Kaplan spent over ten years consulting in the medical device and services space, working with venture, public and private equity. In 2013 he cofounded the technology company Everseat, providing a novel, mobile platform optimizing scheduling efficiency for medical practices, hospitals and health systems. He was a consultant for several years for Cochlear before becoming the Global Head of Clinical Innovation, focusing on modernizing the delivery mechanism and creating the standard of care for cochlear implantation.
He currently resides in Baltimore, Maryland, USA, with his wife and three children.
Uchan, S. February 8th, 2018, blogpost “AI in Medicine – A Historical Perspective” https://rowanalytics.com/blog-post/ai-in-medicine-a-historical-perspective/
Welch, A. May, 29, 2018, CBS NEWS “AI better than dermatologists at detecting skin cancer, study finds” https://www.cbsnews.com/news/ai-better-than-dermatologists-at-detecting-skin-cancer-study-finds/
Taylor, Phil April 13th, 2018, PMLive “Artificial intelligence in healthcare ‘set for 40% growth to 2024’” https://www.pmlive.com/pharma_news/artificial_intelligence_in_healthcare_set_for_40_growth_to_2024_1231723?es_p=6544595