Akron Children’s Hospital

Kevin Watson

Emerging technologies have permeated all aspects of healthcare. Pediatric gastroenterology is not unique among the medical subspecialties, relying on these technological advances to further medical treatment and knowledge. Being board-certified in pediatric gastroenterology and clinical informatics has provided me with a unique perspective on the intersection between medicine and technology. Rapid technological advances have molded how we should think, plan, educate and treat patients in the already technical specialty of pediatric gastroenterology.

Children are not just ‘little adults.’ This is something routinely stated in pediatrics. Adult medicine is often seen as being at the forefront of medicine and technology. Much of what is used in the treatment and diagnosis of pediatric gastroenterology is often extrapolated from experiences seen in adults. To be fair, adults often have higher comorbidities and disease burden and thus more opportunities to assess effects on workflow. In both pediatrics and adults, gastroenterologists perform procedures such as esophagogastroduodenoscopy and colonoscopy, which require varied technological tools and know-how to perform adequately. Pediatric gastroenterology has the unique challenge of patients with smaller anatomy and differential physiology depending on age. Tools that are extremely adequate and efficient for an 18-year-old patient are inadequate for a 2-monthold. With the advancement of ultrathin pediatric gastroscopes, evaluation and treatment of pediatric patients less than 5kg are now possible, including with the ever-growing field of transnasal endoscopy. However, no area of technology has the greatest opportunity for impact on the field of pediatric gastroenterology than artificial intelligence.

Artificial intelligence (AI) is a broad term that encompasses subdisciplines such as machine learning and deep learning. Opportunities in healthcare for AI have heightened with increased computing capabilities and increased data availability. In the endoscopy suite, AI has already been shown to be useful in colonoscopy using computer-aided detection (CAD) for polyps and adenoma detection. This deep learning algorithm acts as a second set of eyes for the endoscopist during the procedure with the goal of improving the adenoma detection rate. Video capsule endoscopy can utilize cognitive neural network (CNN) models to identify potential bleeding lesions, erosions, or ulcers. In inflammatory bowel disease (IBD), endoscopic and histologic evaluation remains the gold standard for investigating mucosal remission. Recent studies show promise in using CAD to assess histologic specimens to identify remission. These technologies not only infiltrate the endoscopy suite but also into other day-to-day maneuverings of the pediatric gastroenterologist.

As a pediatric gastroenterologist and clinical informaticist, I am very invested in how providers and clinical staff interact with the electronic health record (EHR). As gastroenterologists, we get to play with ‘toys’ daily with our procedures, but ultimately, this information obtained will need to be easily retrieved, entered, transferred, or assessed in the EHR. If this process does not occur smoothly, then the patient and provider suffer. Interoperability within and among various healthcare systems is an integral part of the Health Information Management Systems Society (HIMSS) electronic medical record adoption model (EMRAM) Stage 7, a goal for all healthcare systems utilizing EHR. There is now the HIMSS adoption model for analytics maturity (AMAM), which has an ultimate Stage 7 goal in which healthcare organizations would use personalized medicine and prescriptive analytics. Most organizations employ descriptive analytics with more demonstrating maturity in predictive analytics. In the era of COVID-19, increasing usage of patient portals and telehealth have given patients more access to health records than ever before. This will aid in developing personalized medicine protocols with greater clinician and patient interaction. Virtual assistants are a new type of AI chatbots that have the ability to produce personalized responses through natural language processing and machine learning protocols that have been shown to have increased empathy compared to their human counterparts.

Researchers also investigate machine learning models to evaluate therapeutic responses to IBD treatments. Additionally, AI can use machine learning models through support vector machines to assess the risk of developing IBD through genomic datasets. With further validation and extrapolation, models such as these could usher in a new era of personalized medicine. Using digital twins to simulate trials of treatment and to anticipate possible outcomes. It enables healthcare providers to ‘trial’ individual treatment and monitor outcomes before the patient receives it. These advances can aid organizations in reaching the goals of developing prescriptive analytics.

Technology and pediatric gastroenterology will continue to become more entwined in the coming years. With increasing computing power and data through uniform EHR use, we can expect to see increasing adaptation and usage of AI in healthcare, from the endoscopy suite to patient clinics to data analysis. It is imperative that healthcare providers embrace this new technology as with the current rate of advancement, incorporating these models and algorithms is expected to only continue to increase.