Efficient workflows, quality assurance and anomaly detection have been our top priorities since founding Sonio. Today, we are reenforcing our commitment to Quality Assurance with the FDA clearance of the third version of Sonio Detect, which expanded the scope of our AI-powered image evaluation.
What's New in Detect v3
Sonio Detect v3 introduces several new capabilities:
- Expanded scope of automatically detected views: it detects 14 T1 views with 88.6% sensitivity and 98.1% specificity; and 40 T2/T3 views with 90.1% sensitivity and 99.3% specificity
- Expanded scope of ISUOG/AIUM guidelines % coverage: Detect v3 now covers up to 65% of ISUOG and 65% AIUM T2/T3 detailed exam protocols without manual annotation, with a target roadmap set to cover 100% by next release
- Expanded scope of automatically detected structures: while maintaining performance in detecting Brain and Thorax/Heart structures at T2/T3, scope has expanded to include 58 anatomical structures (15 at T1 and 43 at T2/T3) over various anatomical regions: brain, face, profile, thorax, heart, and placenta
- Automatic localization of specific structures on the image to provide users more context on what AI detected
- As with with Detect v2, Detect v3 continues to automatically detect placenta location and fetus sex
The integration of Sonio Detect V3 into our platform is currently underway. While we're excited to bring these enhanced capabilities to our users in the coming months, we're committed to ensuring a seamless rollout and will share specific timing as soon as it's confirmed.
Empowering Quality Care Through AI-Powered Ultrasound
Sonio is an advanced AI-powered ultrasound reporting solution that standardizes fetal ultrasound quality through AIUM or site-specific templates, automated reporting, and intelligent quality checks. With customizable protocols, finding-based reporting templates, and clinical decision support features, the platform empowers healthcare providers to deliver consistent, high-quality care.
Sonio gives both physicians and sonographers confidence in their daily practice while supporting liability and risk management. Sonio is designed to support practices in efforts to reduce avoidable errors, align exams with protocols, improve report coherence for smoother review, and reduce inter-sonographer variability. By automating repetitive tasks such as data entry and report generation, the solution is built to streamline workflows and free up valuable time for patient care, while providing objective signals to upskill sonographers over time.
Our ambition is to bridge the gap between structure, view detection, and anomaly detection to improve both screening and diagnosis in the near future.