Echo labs not using AI will be left behind

 

Echocardiography and artificial intelligence (AI) expert Roberto Lang, MD, director of cardiovascular imaging at the University of Chicago and a past president of the American Society of Echocardiography (ASE), says AI is going to be such a big enhancement to cardiac ultrasound that echo labs not using it will be less efficient, less accurate and see fewer patients than labs that do use it.

Lang spoke in several sessions about ongoing research and development of AI in cardiac ultrasound during the ASE 2023 annual meeting. He also spoke with Cardiovascular Business in two interviews about real-world AI that is either in the final stage of development or already being commercialized.  

"The first thing I will tell you is that in the near future, echo labs that do not have AI will fall way behind. AI is something that will lives inside an echo machine. As soon as a study is done, it will be analyzed by the AI and the reader will get all the measurements. There are already many companies that are very close to having this, and I think it will change echo dramatically," explained Lang. 

Lang said the adoption of AI has been slow until now. In the next couple years, he feels that will change rapidly as clinical evidence mounts to show AI can improve diagnostic accuracy and greatly speed echo workflows. 

"This has been happening slowly, but there will be an explosion in the next couple of years where everybody will be using AI. I don't think you can give yourself the opportunity to have a high-volume lab and not use AI. You are going to have misdiagnoses and you are not going to be very unreliable on the measurements," Lang explained.  

How does AI make cardiac ultrasound more accurate and improve workflow?

Lang said commercialized AI algorithms include a lot of ways to help automated the time-consuming aspects of the echo exams and post-processing. This has grown in importance as echo labs today across the country face a growing sonographer shortage. Lang said time-saving tools can now automatically identify and contour the cardiac anatomy so the sonographer or echocardiographer does not have to spend time drawing lines on the images. This, combined with the AI picking out the textbook example images on which to perform automated measurements, can help standardize quantification for things such as ejection fraction (EF) and strain. The AI also can take a full-volume 3D data set and slice the required views for an echo exam and perform the measurements as well.

He said this will help standardize how measurements are performed and enable reproducibility, regardless of who performed the exam or their level of experience. This also will help make more accurate and consistent measurements across serial exams of patients, which will help better pinpoint disease progress earlier, rather than assuming there is just some variability in measurements from the last person to scan the patient. 

"People will say 'I won't have a job now,' but I think echo people will just have a different type of job," Lang explained. "Instead of people doing this repeated, boring sort of measurements, we will be able to use our brain to do different types of things."

He said the first step in AI was to perfect anatomical identification and contouring of that anatomy. That technology received FDA clearance and has now been on the market for several years. Once you have that and it can be trusted, Lange said the next step was to use the automatically contoured anatomy to allow the AI to pick out the ideal frame and positions in the anatomy to make automated measurements. He said this technology has been coming online commercially over the past few years. Now, Lang explained, vendors are taking things to the next step where AI can look at measurements, determine if they are normal or abnormal and make suggestions for diagnosis. 

He said these steps have all happened sequentially. Many new echo systems that hit the market in the near future are expected to have these capabilities. 

"In the future, you will do an echo and it will come to you with all the measurements, and you will be able to accept or change these measurements." Lang explained. 

Lang also said that this should go beyond standard echo assessments for EF and include more advanced algorithms that assess patients for heart failure, amyloidosis, hypertrophic cardiomyopathy and other, more complex conditions. While physicians can come to the same conclusion as the AI, they can do so much faster if the AI does a first-pass read and offers evidence as to why it thinks a patient has a specific disease state. It is then up to the physicians to double check the AI and use their clinical acumen to decided if the machine is correct, or if there are other factors that need to be considered. 

AI may help identify the types of amyloid patients have

AI is also being developed to able to help classify different subtypes of disease, which is not always that easy for echocardiography today. Often a disease might be detected on echo, but differentiating what subtype it is can sometimes be unclear, or require more advanced imaging. 

One example of this is with amyloidosis, where the subtype of the disease determines the course of treatment. 

"Amyloidosis is an under-detected disease. If I tell you, 'this is amyloidosis,' then you can figure out the type it is. But I think if you wait one or two years, AI will be able to separate the two different types of amyloidosis," Lang said. 

Find more cardiac amyloidosis news and video.

Dave Fornell is a digital editor with Cardiovascular Business and Radiology Business magazines. He has been covering healthcare for more than 16 years.

Dave Fornell has covered healthcare for more than 17 years, with a focus in cardiology and radiology. Fornell is a 5-time winner of a Jesse H. Neal Award, the most prestigious editorial honors in the field of specialized journalism. The wins included best technical content, best use of social media and best COVID-19 coverage. Fornell was also a three-time Neal finalist for best range of work by a single author. He produces more than 100 editorial videos each year, most of them interviews with key opinion leaders in medicine. He also writes technical articles, covers key trends, conducts video hospital site visits, and is very involved with social media. E-mail: dfornell@innovatehealthcare.com

Around the web

Ron Blankstein, MD, professor of radiology, Harvard Medical School, explains the use of artificial intelligence to detect heart disease in non-cardiac CT exams.

Eleven medical societies have signed on to a consensus statement aimed at standardizing imaging for suspected cardiovascular infections.

Kate Hanneman, MD, explains why many vendors and hospitals want to lower radiology's impact on the environment. "Taking steps to reduce the carbon footprint in healthcare isn’t just an opportunity," she said. "It’s also a responsibility."