High-quality data necessary for AI to succeed in healthcare

For artificial intelligence (AI) to develop within healthcare, accessibility to “high-quality data” is crucial, according to a report commissioned by the Office of the National Coordinator for Health IT (ONC) and the Agency for Healthcare Research and Quality (AHRQ).

The report provided an in-depth analysis presented by the JASON Defense Advisory Panel (JASON)—an independent group of scientists and academics that has been advising the federal government, namely the U.S. Department of Health and Human Services—on how AI can assist in improving health and healthcare.

“AI algorithms based on high quality training sets have already demonstrated performance for medical image analysis at the level of the medical capability that is captured in their training data,” the report stated.

For the continued development and “ultimate implementation” of AI in the healthcare setting, JASON recommended that high-quality data be available and accessible for algorithms.

However, those algorithms should not and cannot be expected to perform at a higher level than the training data.

JASON made some recommendations to ensure that high quality data are available for implementation in AI, including:

  • There should be ongoing support for obtaining access to research databases of labeled and unlabeled health data and support for investigations on how to incentivize the sharing of data.
  • There should be ways to identify and develop strategies to fill important data gaps for health data as AI performs poorly if significant data streams are incomplete.

“AI is beginning to play a growing role in transformative changes now underway in both health and health care, in and out of the clinical setting,” the report stated. “At present the extent of the opportunities and limitations is just being explored.”