How AI can make echo labs more efficient
The biggest trend in cardiac ultrasound is the movement toward artificial intelligence (AI) adoption to automate workflows and make them more efficient. While there is apprehension by some clinicians about AI, numerous experts at the recent American Society of Echocardiography (ASE) 2023 meeting said the change is coming and will rapidly revolutionize echocardiography.
"What I want to make sure people understand is that although AI is going to largely replace what we are currently doing today, we must assure that human intuition remains at the helm of clinical decision making and to avoid the assumptions that the machines are always right," explained Steven Lester, MD, a consultant with the department of cardiology at Mayo Clinic College of Medicine and Science, medical director of Discovery Oasis, and founder and chief medical officer of the Mayo Clinic and ASU MedTech Accelerator.
Lester spoke in multiple AI sessions at ASE and took time at the ASE meeting to speak with Cardiovascular Business about this ongoing transformation. Lester said AI is already used everyday in our lives in the form or weather apps, Google Maps, search engines, etc.
"Something truly transformational is happening with AI. We are already seeing how AI is reshaping our modern lives, transforming industries, and it is going to impact probably every aspect of healthcare," Lester said.
Lester compared the worries some people have about AI today to how people felt in the late 1800s when the automation of machinery was seen as taking away the jobs of workers. While that automation did take some jobs, new, high-level jobs were created to manage the machines and in the service sectors. He said a similar transformation will come from AI, which will not replace people in healthcare, but will help make up for the the shortage of clinicians by augmenting the existing humans in clinics and hospitals.
"For any innovation in echocardiography to be widely accepted or adopted, it has to be equalled or exceeded by an innovation in workflow. We are all pressured to do more and more everyday, so we can't add things on, we need to improve the workflow," Lester said.
This will translate into AI ensuring that all the images needed for an echo exam are acquired, for example. The AI also will be able to identify the anatomy, so it can give the echocardiographer all the images from a study on the left or right ventricles, or for specific valves for evaluation. Automated image analytics where the AI will measure everything in an image from textbook landmarks will also make quantification much faster and more reproducible.
"What I care about is, every time I bring this patient back and do the analysis, if an ejection fraction is 56, 56, 56, and it is always measuring the same thing, I can then serially evaluate the patient and see any changes. I think the precision from AI is important. And we have seen cloud-based AI analytics with almost no variability in terms of ejection fraction and strain, as opposed to if you are measuring it, I am measuring it, or another individual is measuring it. Even in the most refined, controlled settings, ejection fractions can be plus or minus 10%," Lester explained.
He added that AI also can help act as a second set of eyes, similar to a second reader, as is used in mammography. AI is now starting to be used at great imaging centers as the second reader on mammography exams to point out anything that the human reader might have overlooked.
He said AI promises to allow physicians more time to spend more time with their patients where the human traits of shared decision making, empathy, judgement and compassion can be used to improve the way medicine is delivered.