Artificial Intelligence and Medicine, What Does the Future Hold?

By: Scott Yerger

     Artificial Intelligence (AI) is already increasing its role in everyday life and we are seeing its rapid evolution right before our eyes.  So, what can we expect when it comes to AI’s role in the future of medicine. In the next 10 years AI is expected to fundamentally reshape medicine, but not by replacing doctors, by augmenting them.

Current Landscape

     On the healthcare administration side, AI will expand its role in driving down cost by automating documentation, improving accuracy in billing and claims filing, streamlining the prior authorization process and better manage scheduling and triage.  Most electronic medical record systems are already looking to embed AI into them.

     On the clinical side AI is already playing a role in assisted surgery and robotics, heart rhythm monitoring, glucose trending and prediction, early signs of infection and even mental health monitoring. AI systems are already matching or exceeding specialists in certain imaging tasks like radiology reviews, retinal disease detection, pathology slide analysis and dermatology disease detection.  Google has AI health models that have already demonstrated high accuracy in disease detection and look to provide faster diagnosis, lower costs associated with screenings and possibly help reduce physician burnout.

Most Impactful Future Roles

     In the next 10 years, AI is positioned to improve quality of life and extend life expectancy in two key areas.  The first will be real time treatment optimization where practitioners will receive real time feedback and guidance with diagnoses through evidenced summarization engines.  This is where the model will measure millions if not billions of clinic data points to compare and guide the practitioner through their objective patient assessments.

     The second and most exciting will be how AI moves medicine from reactive care to predictive care.  The healthcare system today is using basic predictive analytic models to score a patient’s risk level for hospitalization and even readmissions.  These models look at patient data including current comorbidities and vital signs to evaluate how likely a patient is to be hospitalized in a measured period, but this evaluation is generalized as it crosses a broad data set. Future predictive models will be far more personalized.

     Through DNA driven AI engines a patient’s specific risk can be identified and a comprehensive proactive care plan can be generated. Furthermore, AI will evaluate the patient’s specific DNA to determine how that patients body utilizes and responds to medications, even going as far to create patient specific drug compounds that will eliminate side effects and maximize efficacy.

The Digital Twin

     The ultimate patient specific predictive model with the digital twin.  This is where AI will create a computational model of a patients that simulates medication response, surgical outcomes, disease progression and even lifestyle impact.  Organizations like Johns Hopkins University are researching patient-specific simulation models that could allow doctors to “test” treatments virtually before applying them.

Future Challenges with AI

     Data privacy and security are already ever present in healthcare and as patient specific data expands, so does the risk of that information being breached.  If this risk is not adequately mitigated you run the risk of patients opting out of these systems due to lack of trust.  Agencies like the U.S. Food and Drug Administration are actively developing AI-specific regulatory frameworks to ensure patient health information is protected.

     Liability also plays a future challenge. One such example is who will be responsible for AI errors during surgical procedures?  Although errors will be significantly reduced, there is no perfect model when it comes to medicine.

Bottom Line

     Taking medicine and moving from a “treating illness” model to a real time “probabilistic risk” approach will be a huge leap in medical evolution, shifting the focus from better diagnoses to early intervention will be move the current paradigm so much that in 10-15 years we may be looking at our current medical practices the same way we look back at medicine during the dark ages.     

About the Author:

Scott Yerger has over 25 years of Executive Healthcare Management and leadership.  He has an extensive background in the areas of Chronic Illness, Integrated and Value Based Care Models, Physician Relations, Healthcare Information Technology and Strategic Planning.  Over the span of his career, he has also been recognized for his leadership, dedication to patient care and commitment to continuous quality improvement.

Thank you for reading: Artificial Intelligence and Medicine, What Does the Future Hold?