AI can help novice sonographers deliver optimal echocardiograms

 

One of the biggest issues in echocardiography is the inconsistency between how sonographers scan patients and make measurements. However, artificial intelligence (AI) might be able to help overcome this variable and make cardiac ultrasound quantification more consistent. This is especially true for novice sonographers using point-of-care ultrasound (POCUS), where AI is being commercialized.  

Intraoperator variability can often cause five different sonographers to come up with five different sets of measurements, depending on their level of experience, the frames they choose and how they measure distances. This inconsistency is so common that it often leads to cardiologists not worrying too much if measurements change a a little between exams, despite the fact that these changes could also signal a progression of disease. 

Roberto Lang, MD, director of cardiovascular imaging and A.J. Carlson Professor of Medicine with the University of Chicago and a past president of the American Society of Echocardiography (ASE), spoke to Cardiovascular Business on this topic at the ASE 2023 meeting. Lang has been involved with the development of this AI technology from two companies, Caption Health and UltraSight. A third company, EchoNous, showed another version of this AI guidance technology on the expo floor. 

Lang said the technology allows anyone to perform an echo exam and get high-quality diagnostic images. This is why the technology is initially being aimed at the POCUS market, so non-sonographers can capture the proper views needed for a cardiologist or the AI to make measurements.

"The vendors call it 'democratization of echo' because they will help make echo more available. You can talk about physicians in clinics using this, but you can also talk about taking echo to remote parts of the world, and either through AI or Zoom, you can get an interpretation of the echo," Lang said. 

He said these AI systems greatly simplify cardiac ultrasound so that anyone can use it.

"It is easy to do. Actually, my grandson can now obtain an echo," Lang said.

The systems show a thumbnail image in the corner the ultrasound screen to show how to position the transducer and arrows for the direction to move it. Colored bars or a bar graph show the user when they are getting close or are already in the optimal imaging window. The AI then prompts the user to hold position, or how to move or sweep the transducer to complete the exam.

"It is something like a pilot when you see a cross where they want to throw a bomb, these systems actually tell you how to move the transducer over these crosses. It is then interesting that the system will do an auto acquisition. And then the system will use AI to make the measurements on that acquisition. It is an unbelievable thing when you see it working," Lang explained. 

Nurses can use this technology in outpatient clinics to quickly assess oncology patients, and even general practitioners with a handheld ultrasound system in their office can use it to perform a quick triage of a patient.

GE Healthcare signed an agreement to acquire Caption Health in February 2023. GE said it plans to integrate the AI guidance technology into its POCUS systems to help make them easier to use for non-sonographers in the point-of-care market. GE said it has longer term plans to incorporate the technology throughout its other ultrasound systems.

UltraSight announced a partnership at ASE 2023 with Echonous to incorporate its AI guidance technology

Will ultrasound replace the stethoscope?

"We all use a stethoscope; it is a symbol of a physician. But how many know how to use it? But if I can use an ultrasound machine and tell you if your heart is big or small, if it is pumping or not, or if there is pericardial effusion or a huge stenosis, that is a lot of information. In the end, I think you will see this overtake the stethoscope," Lang said.

This prediction was first made by some cardiologists several years ago when small, handheld ultrasound systems first appeared on the market. However, Lang said they were very expensive compared to the cost of a stethoscope at the time. There also are knowledge limitations with non-sonographers that do not have training in ultrasound and not knowing how to perform exams, what to look for, or being used to understand the anatomy in an ultrasound view. 

"Cost was a limiting factor, but the costs have dramatically come down, and with all the AI enhancements, now maybe these systems have a chance of becoming popular," he said.

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

Several key trends were evident at the Radiological Society of North America 2024 meeting, including new CT and MR technology and evolving adoption of artificial intelligence.

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.