Kaiser Permanente Launches Grant Program for AI, ML in Healthcare

The Permanente Medical Group launched the Kaiser Permanente Northern California Division of Research (KP-DOR) Augmented Intelligence in Medicine and Healthcare Initiative (AIM-HI) Coordinating Center, which has committed $3 million to fund health systems’ efforts to use artificial intelligence (AI) and machine learning (ML) to improve diagnoses and patient outcomes.

The program, which will focus on national research initiatives evaluating the implementation of AI and ML to advance diagnostic decision-making, will award grants of up to $750,000 to three to five United States-based health systems to support these efforts.

The press release indicates that AI and ML tools for healthcare have shown some promise—particularly in the realms of medical record notetaking, early warning systems for patient deterioration, and computer vision-based imaging analytics— but AIM-HI seeks to further demonstrate the utility of these technologies in the clinical setting.

"We're aiming to cut through the buzz around AI in health care to prove the promise and positive impact of this exciting technology for improving patient outcomes," said Vincent Liu, MD, principal investigator of AIM-HI, in the press release. "In addition to supporting algorithmic research, the AIM-HI program will develop best practices, and improve the capacity for AI/ML deployment in diverse [healthcare] settings. It's vital that we use these powerful tools thoughtfully and in a way that scales to benefit patients, physicians and care teams."

AIM-HI’s request for project proposals is now open, and letters of intent are due June 30, 2023.

Proposals will be assessed by a national advisory committee and expert reviewers from Kaiser Permanente, the press release notes. Selected project teams will work closely with the committee and reviewers over the 2-year grant period to leverage Kaiser Permanente’s experience in health AI.

"At Kaiser Permanente, we've demonstrated the real-world benefits of AI and ML with groundbreaking algorithms," said Stephen Parodi, MD, executive vice president of The Permanente Federation. "Our early warning systems are at work 24 hours a day, 7 days a week to identify high-risk hospitalized patients. One of these programs has saved an estimated 500 lives a year. With this initiative we hope to support the spread of impactful work in health systems of all sizes across the country."

This announcement builds on Kaiser Permanente’s ongoing work on healthcare AI, ML, and other advanced technologies.

In 2021, the health system shared the results of a study demonstrating the utility of natural language processing (NLP) technology to assist with the diagnosis of aortic stenosis, a common type of heart disease.

The researchers noted that identifying complex conditions and disease trends can be difficult because health systems often track patients using procedure or diagnosis codes, which are mostly used for billing and therefore not particularly useful for more specific research or clinical care.

To address this, the research team trained an NLP tool to sift through patient EMRs and echocardiogram reports to flag abbreviations, words, and phrases typically associated with aortic stenosis. Doing so allowed the researchers to identify thousands of patients with the condition.