Image by Shapelined

Distilling the Data into Useful Information: The Rise of Artificial Intelligence

Updated: May 7

in the Practice of Pediatric Cardiology

Addison S. Gearhart, MD and Anthony C. Chang, MD, MBA, MPH, MS


What does the future of Artificial Intelligence (AI) in Pediatric Cardiology hold?

Instantaneous access to relevant, precise, and predicative insights about each patient before the clinic visit. Data from that patient– healthcare records, personal fitness trackers, vital signs, social history, Electronic Medical Records (EMR) etc.– autonomously added into a collaborative database that syncs, interprets, and translates each patient’s health profile against a large aggregate of shared peer health data and outcomes, the current medical body of literature, and ongoing trials (Figure 1) to enhance patient care and elevate the quality of discussion during the face-to-face patient encounter. Useful information from this data could include: morbidity and mortality prediction, drug therapy suggestions targeted to each patient’s genetic profile, tailored surgical plans, recommendations for timing of sequential imaging, and autonomously-read advanced imaging studies. The other arm of emerging AI systems, virtual assistants and robotic process automation (RPA), could ease the after-clinic responsibility workload by offering 24/7 assistance to look up patient records, search specific medical questions, schedule appointments, write notes, complete orders dictated in the exam room, check on a patient’s progress and even answer some of the patient’s basic medical questions.

FIGURE 1 A dramatization of a future AI healthcare interactive platform. [41]

In our current model, all too often wasted data sits on the wrists of many of our patients, in popularized genetic testing databases, and in EMR. Data is piling up urging the clever physician to find ways to gain access while protecting privacy and for the even cleverer data scientist to develop smarter algorithms capable of distilling the heterogenous, often incomplete forms of medical data, into clinically meaningful and actionable information (Figure 2). Those committed to advancing the field are fueled by this limitless potential to revolutionize the cardiologist’s tool set by replacing the current labor-intensive task of finding meaningful connections in an ever-expanding sea of data with interpreting and acting on data-driven insights. Proponents argue this prototypical AI system will usher in an era of unparalleled medical intelligence, redefining the patient’s path toward better health.

It is increasingly apparent that with the sustained rate in the rise of publications and advances in AI in and outside of the medical field that we face an inescapable and exciting future in which AI will influence how we as physicians practice medicine and how our patients receive healthcare. The question of if there will be a future partnership between AI and pediatric cardiology is quickly being replaced with the more pressing questions of why now, how, who and when. In Part I of this review article we will address the questions of why now and how and in Part II we will discuss the questions of who and when.

To read the full article, please go to the January 2019 Issue of CCT.