The potential for artificial intelligence (AI)to reshape the field of health care — to help improve diagnosis and enable an increasingly precision approach to medicine — may seem boundless. But the lack of technological infrastructure, nascent policy debates, and public distrust around AI are likely to hinder its ascent, according to a new paper from Harvard T.H. Chan School of Public Health researchers.
The article, published in NPJ Digital Medicine on August 16, 2019, provided a detailed analysis of why the AI algorithms that feature prominently in health care research literature are not of much use at the frontlines of clinical practice at present. Among the key challenges is that most health care organizations do not have the data infrastructure required to harness the huge amounts of data that are needed to refine and improve AI algorithms.
The paper noted that health systems presently have two options when it comes to AI: they can reign in the enthusiasm regarding the potential of AI in everyday clinical practice or they can tackle issues such as data ownership and public trust around AI and then make significant investments into the necessary infrastructure to fully optimize their use of AI.
In an accompanying blog post, study co-author Trishan Panch, instructor at Harvard Chan School, wrote, “Like all public health interventions, [AI] has the potential to create enduring benefit but will require not just a broad coalition of support and partnership between the public and private sector but also the trust and enduring support of patients.”
Harvard Chan School’s Heather Mattie, instructor of data science and executive director of the master’s program in health data science, was also a co-author of the paper.
Read the NPJ Digital Medicine article: The “inconvenient truth” about AI in healthcare