Despite the regular publication of scientific studies demonstrating the myriad life-saving abilities of artificial intelligence, the technology will not translate to the actual practice of healthcare without a stronger data infrastructure.
In a new perspective article in npj Digital Medicine, a group of medical researchers and data scientists describe this "inconvenient truth": that AI will not have a widespread effect on healthcare until organizations have not only developed appropriate methods of storing and computing data, but also devised ways to share that data with one another.
This will take significant upfront investment from organizations, which, for the most part, have yet to find a compelling reason to do so, according to the authors. External support is lacking, too: increased interoperability would decrease the power of EHR vendors, and while government policies surrounding data privacy are "well intentioned," per the article's authors, they are unsustainable without financial backing.
Without taking the necessary steps to rectify these major concerns, they wrote, "opportunities for AI in healthcare will remain just that — opportunities."
Read the full article here.
More articles on AI:
Google DeepMind co-founder placed on leave for alleged controversial projects
Tampa General opens high-tech clinical command center
AI pilot project to help CMS identify fraud deemed a success