ASU's role in the future of AI in Healthcare

3D Holography, AI and Smart Medical Imaging and X-rays for Doctors http://www.brother.co.uk/business-solutions/healthcare/future-of-hospital-technology

Last week I attended a Pop-Up conference on the Future of AI in Healthcare presented by Arizona State University's (ASU) Health Entrepreneur Accelerator Lab (HEALab) and Systems Imagination Inc.  It was awesome to see how Arizona is coming along as a hub for entrepreneurship, particularly at the intersection of bioscience and artificial intelligence (AI).  The pop-up presentations provided valuable insight into four areas of consideration when it comes to the delivery of value with AI particularly in healthcare:  Technology, Politics & Policy, Social Acceptance, and Organizations.  

Technology

Aural Analytics is an Arizona start-up that uses AI to provide objective measures of brain health.  This patient-focused application received the 2017 Scrip Award for the best technological development in clinical trials.  The AI engine searches for subtle, but clinically relevant changes in brain health.  This is a super cool use of AI that is already producing results for clinical trials and neurological research.   And, according to a company representative, this is a company that is already turning a profit.   Hybridchart  is a Software as a Service (SaaS) application designed by an Arizona physician that offers to solve the issue of multiple disconnected electronic medical records within a single care provider entity.  Both companies presented solutions that provide compelling opportunities for improving patient experience, improving population health, and reducing the cost of care.  

Politics and Policy

Gary Marchant, Lincoln Professor of Law, Ethics, and Technology and Faculty Director of Science & Innovation at the Sandra Day O'Connor College of Law gave us insight into the litigious nature of technology.  Gary pointed out to us that the public holds technology to a higher standard of quality than humans.  He provided an example of how a self-driving car that may, in fact, be safer and have fewer accidents than a human-driven vehicle is more likely to be sued when something goes wrong.  George Runger, Chair of Departments of Biomedical Informatics and Biomedical Diagnostics, pointed out that the FDA has greater expectations of technology solutions than existing physician standard of care.  Case in point:  The FDA required accuracy for skin cancer recognition needs to be at the level of accuracy that a dermatologist will provide or higher.  All of this slows down the adoption of new technology in healthcare.  With AI, models need to be developed using real data to improve performance.  The challenge is how do you get sufficient data to improve accuracy and yet pass public and FDA scrutiny while the AI algorithm is learning?

Another eye-opener, at least for me, was the prominent role that local government plays in healthcare delivery.   According to the National Conference of State Legislatures, our 50 state legislatures make thousands of health policy decisions every year (Ballotpedia.org).  The decisions include not only budget appropriations, but also; which services are covered by insurance, and how personal healthcare information is managed.  Heather Carter, Arizona House of Representatives, Clinical Assistant Professor, ASU College of Nursing and Health Innovation, provided an example of recently passed bill H.B. 2042 that expanded insurance coverage requirements delivered through telemedicine.  A bill to extend telemedicine coverage to urology failed to pass for 4 years in a row until it got written into the H.B. 2042.  A bill addressing the recent opiate hot topic of pain medicine and substance abuse.  Heather's point was that entrepreneurs need to consider policy and politics early in their strategy development.  And be prepared to communicate the value of the technology to local politicians and their constituents.  

Social Acceptance

Woven in throughout the panel discussions was the subject of privacy and security.   The public demands innovation in healthcare, but naturally, privacy is a concern.  And public concern informs policy and motivates policymakers.  For Artificial intelligence applications to effective, patients and consumers need to be willing to share data.  Otherwise, the application's algorithms can't learn and become more and more accurate.  New privacy legislation will likely be put forth to address the concerns elevated by the Facebook and Cambridge Analytica scandal.  The spillover into the healthcare domain and how new privacy and security policies will impact the advancement of AI solutions is still to be seen.

Organizations
​
Frequently forgotten is the importance of the healthcare organizations that need to not only adopt new technologies but also adapt to them.  Silo's within a healthcare organization impede the delivery of value especially when new technology is implemented without consideration for workflow and organizational structure.  According to the consulting firm McKinsey, executives tend to focus too much on individual initiatives and not on how the business must change.  McKinsey suggests that in the digital world, in order to think holistically organizations should organize around customer journeys.

This year's pop-up event presented by ASU's HEALab and Systems Imagination, Inc, provided not only a view into the Artificial Intelligence work of the many schools and colleges of ASU, but also excellent insights into social, political and organizational considerations for the delivery of  entrepreneurial success and value back to our community of healthcare providers and consumers.  I look forward to seeing ASU's influence on technology and healthcare at HEALab's next event!

---