The AI algorithm delivered a 22-point improvement in anomaly detection in a multicenter study. 

Sonio, a cloud-based ultrasound reporting and image management solution provider, announced that the US Food and Drug Administration (FDA) has granted clearance for its latest artificial intelligence (AI) module, Sonio Suspect, designed to improve prenatal diagnostics by enhancing the detection of fetal anomalies.

A multicenter reader performance study involving 47 sites, including 37 in the US, demonstrated a statistically significant 22-point improvement in anomaly detection (from 69% to 91% AUC, p<0.001). This improvement was confirmed across diverse patient demographics, including BMI, and was consistent regardless of clinician background or experience—whether maternal-fetal medicine specialists, obstetricians and gynecologists, or radiologists.

Additionally, Sonio Suspect enables early identification of subtle congenital malformations from as early as 11 weeks. By providing a broader diagnostic window, it supports clinicians in identifying abnormalities sooner, giving families and healthcare providers more time to act, plan interventions, and improve both maternal and fetal outcomes. 

[RELATED: FDA Clears Clarius Mobile Health OB AI Fetal Biometric Tool]

“The clearance of Sonio Suspect is a major milestone for Sonio, at the core of our mission to improve patient outcomes in maternal-fetal medicine,” says CEO Cécile Brosset, in a release. “By combining real-time AI quality control with AI-driven anomaly detection, Sonio supports ultrasound providers at every step of the patient pathway, from exhaustive documentation to accurate diagnosis. Our technology is designed to help healthcare providers detect issues early and streamline processes, ultimately improving the care every patient receives.”

Reducing Missed Fetal Anomalies

Research shows that up to 51% of fetal anomalies are missed during standard prenatal ultrasound screenings, with 31% of these missed cases resulting from misinterpretation of high-quality images. Sonio Suspect aims to tackle this challenge by providing automatic detection of eight abnormal findings across seven ultrasound views of three key fetal anatomical regions: the heart, brain, and abdomen.

Sonio Suspect will be integrated into Sonio’s reporting and workflow solution, which includes Sonio Detect, a quality assurance algorithm designed to ensure high-quality image acquisition.

Photo caption: Sonio software

Photo credit: Sonio

Summary:

The FDA has cleared Sonio Suspect, an AI-powered ultrasound module designed to improve prenatal diagnostics by assisting in the detection of fetal anomalies. A multicenter study across 47 sites, including 37 in the US, demonstrated that Sonio Suspect improved anomaly detection accuracy by 22 percentage points, achieving an AUC of 91%. The AI tool enables earlier identification of subtle congenital malformations as early as 11 weeks. It will be integrated into Sonio’s reporting and workflow solution alongside Sonio Detect.

Key Takeaways:

  • FDA Clearance for AI-Powered Detection – Sonio Suspect has received FDA clearance for assisting in the detection of fetal anomalies in prenatal ultrasound.
  • Proven Improvement in Anomaly Detection – A multicenter study found Sonio Suspect increased anomaly detection accuracy from 69% to 91% AUC, with consistent results across clinician experience levels and patient demographics.
  • Early Identification of Fetal Abnormalities – Sonio Suspect can detect congenital malformations as early as 11 weeks, providing a longer diagnostic window for early intervention and care planning.