AI adoption for medical imaging soars

More than half of survey respondents reported using an AI algorithm for at least one imaging use case, compared with just 17% in 2018.

Article By: Emily Olsen

Blog Source From : https://www.healthcaredive.com/

Dive Brief:

• The number of healthcare organizations using artificial intelligence in medical imaging has ballooned over the past six years, according to a report published last week by Klas Research. 
• More than half of survey respondents have started to utilize AI algorithms for at least one imaging use case, the healthcare IT research firm found. In comparison, only 17% of healthcare organizations reported piloting or actively using an AI tool in 2018.
• About three-quarters of healthcare organizations that are currently considering an AI purchase have specific tools and use cases in mind — suggesting “strong momentum and future uptake,” according to Klas.

Dive Insight:

AI has become an appealing new tool for the healthcare industry, with proponents arguing the products could help alleviate providers’ heavy workloads — a major concern for health systems managing clinician shortages and burnout. 

The technology has already been integrated into a growing number of medical devices. Last year, the Food and Drug Administration authorized 221 AI-backed medical devices, compared with just six in 2015, according to an analysis by MedTech Dive.

More than three-quarters of those AI-enabled devices have been for radiology use cases, like flagging potential health conditions or planning for radiation therapy.

The latest survey by Klas, which interviewed more than 200 health systems and imaging groups about their use of the technology, found adoption of AI imaging tools has soared as the FDA greenlights more products. 

More than half of healthcare organizations are using an AI imaging tool

AI use across the imaging market by live tools and future adoption plans

Companies with large imaging volumes are more likely to have an AI product currently live. But more than 30% of midsize and small organizations said they plan to adopt the technology for the first time soon.

Respondents reported using 65 different commercial AI products — from companies like RapidAI, Viz.ai and Aidoc — as well as homegrown tools. 

The most common use is for neurology and stroke imaging, such as prioritizing the most urgent patients, according to the Klas. But healthcare organizations also adopted the tools to identify lesions in mammograms, detect lung nodules or automatically generate reports for radiologists to review and edit.

However, most organizations are employing the products for few use cases. Nearly 75% of respondents said their companies only have one or two use cases, and fewer than 10% have five or more.