Fetal ultrasound (US) is the primary imaging modality for prenatal screening, yet its interpretation relies heavily on the expertise of the clinician. Despite advances in deep learning and foundation models, existing automated tools for fetal US analysis struggle to balance task-specific accuracy with the whole-process versatility required to support end-to-end clinical workflows. To address these limitations, we propose FetalAgents, the first multi-agent system for comprehensive fetal US analysis. Through a lightweight, agentic coordination framework, FetalAgents dynamically orchestrates specialized vision experts to maximize performance across diagnosis, measurement, and segmentation. Furthermore, FetalAgents advances beyond static image analysis by supporting end-to-end video stream summarization, where keyframes are automatically identified across multiple anatomical planes, analyzed by coordinated experts, and synthesized with patient metadata into a structured clinical report. Extensive multi-center external evaluations across eight clinical tasks demonstrate that FetalAgents consistently delivers the most robust and accurate performance when compared against specialized models and multimodal large language models (MLLMs), ultimately providing an auditable, workflow-aligned solution for fetal ultrasound analysis and reporting.
Alex: Welcome to another episode of ResearchPod. Sam, I've been thinking about prenatal checkups—how doctors use ultrasound to check on babies before they're born, but there just aren't enough trained people to do these scans properly everywhere. What's the latest research doing to help with that?
Sam: This paper introduces FetalAgents, a system designed to handle fetal ultrasound analysis more like a team of specialists working together. Fetal ultrasound uses sound waves bounced off the baby to create pictures of its growth and health during pregnancy—it's the main way doctors check for things like size and development. The core puzzle is that sonographers—the trained technicians—have to manually spot the right views of the baby, measure parts like the head or belly, and note everything down, often from shaky video streams full of noise from the equipment. That's error-prone and slow, especially in places short on staff.
Alex: So this is tackling the shortage of ultrasound experts by automating the whole workflow? Like turning a video scan into a full medical report without a human having to switch tools constantly?
Sam: Yes, exactly. The paper builds FetalAgents to mimic that human process end-to-end.
Alex: Right, that makes sense—the bottleneck is coordinating all those steps, not just seeing the images. Walk me through how this team-of-specialists idea works in practice.
Sam: Picture a busy hospital where a triage nurse looks at a patient, figures out the issue, and sends them to the right doctors—like a heart specialist or bone expert—who then report back for a full summary. In FetalAgents, a central coordinator AI does that: it gets an ultrasound image or video plus a question like "What's the baby's head size?", identifies the body part shown, and sends it to the best-suited specialist AIs for things like spotting the view or measuring. Researchers call this a multi-agent system, but it's really about smart teamwork to avoid mistakes from one AI guessing everything. The specialists use proven tools trained on thousands of scans, and they combine results in fixed ways to stay reliable—no wild guesses.
Alex: Okay, so the coordinator is like that triage nurse directing traffic. But why does splitting into specialists beat using one big AI brain for the whole thing?
Sam: Single AIs shine at one job, say measuring the head circumference—which is the distance around the baby's skull to track growth—but falter when jumping between spotting views, outlining organs, or handling full videos. By routing dynamically, FetalAgents picks the strongest tool per step. This setup emulates sonographers switching focus across a scan, leading to more consistent outputs structured as clean data lists, which a final summarizer turns into reports.
Alex: So it's not just better at individual bits, but the coordination makes the whole chain stronger. What kinds of specialist tools are they pulling from?
Sam: The experts wrap specific vision models—AI systems tuned to "see" ultrasound patterns. For plane classification, they combine models like one that understands ultrasound language from vast training and others for fine details. Segmentation experts outline structures like the abdomen. Biometry pulls measurements, like abdominal circumference via ellipse fits on those outlines. All outputs stay in structured formats to keep things auditable and hallucination-free—just numbers and facts.
Alex: Got it—each expert handles a narrow skill, like outlining the stomach or estimating gestational age, which is how far along the pregnancy is from measurements. But real scans are videos, not still photos—how does it deal with that continuous stream?
Sam: Videos are sequences of frames from the probe moving over the belly, so sonographers pick key moments manually. FetalAgents automates that: first, a tuned model spots diagnostic keyframes across planes like femur or thorax. Then the coordinator analyzes each, runs experts frame-by-frame, and the summarizer aggregates into a report with patient details, like checking measurements against growth charts. This mirrors the full workflow, from raw video to typed notes with consistency flags.
Alex: Oh—so that's why it's workflow-aligned; it handles the messiness of real exams, not just perfect lab images. And the paper tests this on actual external data, right?
Sam: Yes, they trained on public datasets then validated on independent multi-center ones, like African clinics, for out-of-distribution robustness. Across eight tasks—plane ID, segmentations, biometry—FetalAgents topped specialized models and large multimodal language models. For instance, on abdomen segmentation, it reached a notably higher Dice score—a measure of outline match—than the next best.
Alex: That's a clear improvement, but does the ensembling always win, or are there catches?
Sam: Ensembling boosts robustness by fusing strengths, like correcting one model's edge slips with another's, but it relies on those pre-trained vision models working well together. The paper notes performance holds on tested tasks, yet may need tweaks for rare anomalies or new protocols—prospective trials in clinics are next.
Alex: Makes sense; it's a meaningful step for standardizing scans where experts are scarce. So zooming out, what could this mean for prenatal care down the line?
Sam: In low-resource areas, it could cut delays by automating reports from any ultrasound video, easing sonographer burdens and spotting issues earlier. The structured outputs aid audits too, aligning with global standards like WHO charts. Still, as the authors stress, real-world deployment needs those clinical trials to confirm safety across diverse cases.
Alex: Well put, Sam. This shows how agent coordination can tackle complex medical imaging without overpromising. Thanks for breaking it down—listeners, check the shownotes for the paper. Thanks for listening to ResearchPod.