Sonio Suspect

Sonio Suspect is a cloud-based software tool that assists physicians during fetal ultrasound examinations by automatically detecting and characterizing abnormal fetal findings (such as cardiac, abdominal, and cephalic abnormalities) from gestational weeks 11–41 using machine learning. The device provides a concurrent reading aid on acquired images during or after the examination, and the user manual explicitly states that patient management decisions should not be made solely on the device output.

Both Sonio Suspect and the predicate Koios DS use computer vision and machine learning-based algorithms to analyze ultrasound images. The main technological difference is that Sonio Suspect is a secure cloud-based and standalone software compatible with ultrasound systems, whereas Koios DS is an ASP.NET web application deployed to a Microsoft IIS server in a Windows environment. Both operate on the same technical principle of image characterization and share the same product code (POK).

Not stated in this summary. The document references 21 CFR §892.2060 special controls but does not cite specific consensus standards such as ISO, IEC, or ASTM.

Sonio Suspect is substantially equivalent because it shares the same intended use (assisting physicians in image analysis using machine learning), similar clinical outcomes (providing diagnostic information), and the same technical principle as Koios DS. Although the target population differs (fetal vs. breast/thyroid), the user population type (interpreting physicians) and clinical application model are comparable. The device demonstrates equivalent safety and effectiveness through standalone performance testing (93.2% sensitivity, 90.8% specificity) and clinical MRMC study showing superior assisted-reading accuracy (21.9 percentage point improvement in AUC). The technological differences in platform deployment do not raise new safety or effectiveness questions when used as labeled.