A Bit of Magic: How This Women’s Health Ultrasound Tool Automates Tasks to Help Clinician Efficiency

GE Healthcare

“The first time I saw an obstetrics ultrasound exam, I thought it was a little bit like magic,” said Dr. Susanne Johnson, Associate Specialist in Gynecology at Princess Anne Hospital in the United Kingdom. “I thought to myself, ‘now that’s something I want to do.’”

Now as a gynecologist who specializes in helping diagnose some of the most difficult gynecological diseases and teaching others her skills, Dr. Johnson is using advanced ultrasound technology that has a new bit of "magic" built-in: automated tools and artificial intelligence.

“Every patient I have is a bit of a puzzle and my job is to try and develop a hypothesis of what’s the problem. Then, I can use ultrasound to try and prove it and point the patient in the right direction to get the care they need,” said Dr. Johnson.

Despite this passion, Dr. Johnson is one of many women’s health clinicians facing workflow and protocol challenges that can make her job more difficult. In fact, a recent study found obstetrics (OB) and gynecology (GYN) clinicians in the United States are reported to have some of the highest burnout rates in the clinical community, attributed to bureaucratic tasks like paperwork and charting.[1]

“I always use the latest research and protocols in my practice, but it can be difficult for clinicians to do so as there are few examples of how to do this,” said Dr. Johnson. “Trying to use a systematic way of scanning and making sure you get good images and measurements of everything you need can actually be quite difficult.”

On top of the typical challenges, women’s health clinicians are now facing additional pressures in today’s pandemic environment.

“We’re being asked to see more patients than ever,” said Dr. Johnson. “In addition to this, we’re trying to get patients in and out of the room as efficiently as possible to limit any sort of possible exposure to COVID-19.”

To help combat these constraints, clinicians are looking for technology to automate tedious tasks allowing them to focus on the examination and their patients.


Industry-First Auto Recognition Tool Improves Clinical Confidence

For expecting moms, the second trimester pregnancy ultrasound exam is used to check the development of the fetus and flag any possible concerns.

When performing this critical second trimester pregnancy ultrasound exam, the International Society of Ultrasound in Obstetrics and Gynecology requires practitioners to capture, label, record and store twenty images of the fetus throughout an exam and ensure all are accessible in the patient’s record for future reference.

In an ideal world, the practitioner performing the exam would have an assistant to help ensure image quality and check each of the required images are captured. However, that isn’t the case for most practices.

It was this thought that inspired Professor Aris Papageorghiou, a clinician scientist in Fetal Medicine who divides his time between clinical work at St George’s Hospital in London and research at the University of Oxford, to help develop SonoLyst**. This is the industry’s first integrated AI algorithm that recognizes the 20 views recommended by the International Society of Ultrasound in Obstetrics and Gynecology mid-trimester practice guidelines for fetal imaging.

The new tool is one of many intelligent applications on the recently released Voluson SWIFT ultrasound system to make the jobs of women’s health practitioners easier.

Animated GIF- SonoLyst Auto Recognition.gif

“If we don’t get the quality of images that we need, it can mean we could have missed something and we need to call the patient back in to be scanned again which is inconvenient for the patient and time consuming for us,” said Prof Papageorghiou. “With a lack of manpower in many clinics, we can use artificial intelligence to serve as a type of assistant and peer reviewer.”

To use SonoLyst, clinicians simply perform the second trimester ultrasound exam as normal. When they freeze the image on the screen, the AI tool automatically identifies anatomy, selects all applicable annotations and measurements, and ensures the image quality meets clinical standards. The operator then confirms the information captured and the data is entered into the scan checklist and report, enhancing workflow and reducing variability between operators for improved consistency.

“It’s really useful for workflow and reducing variability between operators. If you’ve got a busy day and scanning patient and after patient, it’s possible that you might forget to check a certain view,” said Prof. Papageorghiou. “Before you end the exam, the system can give you a reminder to say ‘hey, you didn’t get the kidney image you need’ or ‘that image of the spine wasn’t of high enough quality.’”

In practice, the new SonoLyst tool can optimize the scan workflow by 73 percent when compared to manual 2D workflow compared to previous systems.


Guided Workflows and Tools Help User Efficiency

These tasks and challenges aren’t unique to second trimester ultrasound exams, and for many users these challenges can be amplified by not using the tool very often or being new to ultrasound and the many intricacies that come with it.  

“While some women’s health practitioners are experts at ultrasound and do it every day, many OBGYNs are only using ultrasound occasionally,” said Prof. Papageorghiou. “Although they get initial training in ultrasound it takes a while for them to really build the skill set in a specialty like women’s health. It’s a bit like getting your driver’s license: you may have the license, but it takes you a while to really get a good feel for it.”

To help, the Voluson SWIFT was designed with fewer buttons and a touch screen with an intuitive user interface, replicating touch gestures you would use on a smartphone like zooming in on an image by pinching your fingers together. In addition to hardware updates, onboard software tools like ScanAssistant guide clinicians through protocols and reduces the patient scanning time by up to 45 percent.[2] A clinician selects the protocol and the system helps ensure all the required images are captured accordingly.

“These tools are a bit like a phone-a-friend IT solution on your ultrasound,” said Prof. Papageorghiou. “It’s good for occasional users, new users, or recently trained users. When using this in a clinical setting, the image quality was great, and the AI worked very well.”

For more-consistent ultrasound users, the system allows each person to customize and create their own user interface on the system.[3],[4] Just like you’d build your own home screen on your iPhone, clinicians can save their favorite settings, protocols and image adjustments.

“I do think one of the exciting things about the Voluson SWIFT is that it expands the access to this level of ultrasound technology, image quality and AI,” said Prof. Papageorghiou. “From my perspective, this is just the beginning of this type of artificial intelligence in women’s health ultrasound, which makes me excited for the future.”

 **SonoLyst incorporates the AI technology of Intelligent Ultrasound

[2] Comparison performed using GE’s Voluson SWIFT/ SWIFT+ vs. Voluson P8.

[3] Comparison performed using GE’s Voluson SWIFT/ SWIFT+ vs. Voluson P8.

[4] Comparison performed using GE’s Voluson SWIFT/ SWIFT+ vs. Voluson P8 BT18 & Voluson S6 BT16.