Generative AI is reshaping healthcare, and its emerging role in ultrasound reporting is gaining attention. To ensure we are aligning innovation with clinical realities, GE HealthCare surveyed a group of North American radiologists who specialize in ultrasound. We asked where they see promise, what benefits they prioritize, and where they still have concerns when it comes to integrating generative AI into ultrasound reporting workflows. Their responses offer a compelling snapshot of how this technology is being viewed today—and where it may be headed tomorrow.
This three-part series explores what we learned. In Part 2, we focus on the benefits radiologists are most eager to see from generative AI.
Top-ranked benefits of generative AI in ultrasound reporting
To pinpoint where generative AI could offer the greatest value, we asked radiologists to rate a range of potential benefits on a scale from 1 (strongly disagree) to 5 (strongly agree). Two priorities stood out:
- improving quality through standardization
- enhancing workflow efficiency
Quality and standardization win out
Across all the survey questions, the highest level of agreement came in response to generative AI’s potential to improve quality and consistency in ultrasound reporting. Every radiologist surveyed agreed—most of them strongly—that one of generative AI’s greatest strengths is its ability to standardize how findings are documented.
Consistency is especially critical in ultrasound, where exams are often highly operator-dependent and interpretation can vary. Variability can lead to delayed follow-ups, miscommunication among providers, unnecessary procedures, and, in some cases, misdiagnosis. Generative AI solutions that are trained on vast ultrasound data sets can provide consistent measurements, recognize patterns, and guide users through standardized documentation workflows. This means they can enforce structured language, ensure all required fields are completed, and align reporting practices with institutional templates offer a powerful remedy to this long-standing challenge.
“When generative AI helps standardize how findings are described and how measurements are presented, it builds trust, not just between radiologists, but also between departments and across the care team,” says Ravi Nagavarapu, Senior Global Product Manager, Digital Ecosystems for Medical Device Customers at GE HealthCare. “It brings everyone onto the same page, which is critical for both quality and continuity of care.”
By improving uniformity in how findings are documented, generative AI can help create a stronger, more reliable foundation for diagnosis, clinical decision-making, and patient care.
Efficiency gains through automation
Next, the radiologists overwhelmingly recognize generative AI’s potential to enhance workflow speed and efficiency. In follow-up discussions, participants noted that generative AI could accelerate the reporting process by automating repetitive tasks and generating pre-filled templates.
“There’s a real opportunity here for generative AI solutions to take on the mundane, time-consuming aspects of reporting,” says Nagavarapu. “Even saving a few minutes per report can add up to hours across a week, and that’s time radiologists can reinvest into more complex cases, teaching, or collaboration.”
Radiologist burnout and boreout remain pressing concerns, amplified by a global radiologist shortage that continues to stretch already limited resources.2,3, This means even modest gains in reporting efficiency via generative AI could have a meaningful impact on radiologist well-being and departmental performance. This potential goes beyond ultrasound— the majority of clinicians across specialties believe it’s very important to use advanced technologies to make every day clinical tasks more efficient.⁴
Mixed responses on accuracy benefits
When it came to improving diagnostic accuracy, however, opinions were far more divided. Responses ranged from strong agreement to neutrality to even strong disagreement. This was the only option with such a broad spectrum of responses.
“Radiologists are cautious, and for good reason,” Nagavarapu commented. “They want tools that enhance their ability to do the job, but they don’t want a black box making interpretive decisions without their input. When it comes to diagnostic accuracy, it’s not just about what the AI says, it’s about understanding how it got there.”
In follow-up discussion, participants cited concerns about how generative AI might handle edge cases or atypical presentations, emphasizing the need for ongoing validation, transparency, and clinician involvement in AI-solution development.
Edge cases sparked another thought-provoking discussion: where can generative AI deliver the most interpretive value? Should it assist with the “bread and butter” cases, meaning straightforward, high-volume studies where AI could help reduce cognitive load and accelerate reporting? Or should it focus on the most complex exams, where radiologists are more likely to need support? At first glance, it may seem ideal to offload routine cases and let AI handle them independently. But there’s a tradeoff. Leaving radiologists with only complex, nuanced cases could increase stress and fatigue over time. As one survey participant put it, “Think about what your day looks like when you're left with only edge cases and all day long, hundreds of times a day.”
Cost savings is not the central focus
Interestingly, cost savings did not emerge as one of the top benefits of applying generative AI to ultrasound reporting. Few radiologists cited it as a major expected benefit, suggesting that technology’s value is being evaluated more through the lens of quality and productivity gains than direct financial returns.
“Cost always matters, but when you’re talking about reporting, radiologists care more about quality and time,” says Nagavarapu. “If generative AI can help reduce repeat exams, improve consistency, and give radiologists some time back, those benefits can drive value in more meaningful ways.”
That said, the group did acknowledge that improved efficiency could translate into broader health system savings. Avoiding unnecessary scans was cited as an example.
There are several data points that could be considered to measure the economic impact of AI-assisted ultrasound workflow. These include cases per day, read time, reporting time, backlog rate, and overtime costs. However, the results of this survey suggest financial benefit is viewed as secondary to clinical and operational improvements.
Where radiologists see immediate value
Among this sample of radiologists, the benefits of generative AI in ultrasound reporting are seen in terms of report consistency, documentation quality, and enhancing workflow efficiency. While questions about accuracy and broader financial impacts remain, the early focus is clear: Generative AI can offer real, tangible support in helping radiologists work faster, more consistently, and with greater confidence in the technical quality of their outputs.
Stay tuned for Part 3 in our series, where we explore the barriers, radiologists see to adopting generative AI.
REFERENCES
1.) Internal data from a GE HealthCare focus group, October 2024
2.) “Radiology Facing a Global Shortage,” RSNA News, last modified May 10 2022, https://www.rsna.org/news/2022/may/global-radiologist-shortage.
3.) “ACR, RBMA Urge Congress to Address Workforce Shortages,” American College of Radiology, last modified March 23 2023, https://www.acr.org/Advocacy-and-Economics/Advocacy-News/Advocacy-News-Issues/In-the-March-25-2023-Issue/ACR-RBMA-Urge-Congress-to-Address-Workforce-Shortages.
4.) 2022 GE HealthCare Reimagining Better Health study. Results on file.
©2025 GE HealthCare. GE is a trademark of General Electric Company used under trademark license.
JB35592XX
