Sonio Suspect

Sonio Suspect is a machine learning-powered web application designed to assist healthcare professionals in identifying and characterizing abnormal findings in fetal ultrasound images. It leverages AI to automatically detect views and evaluate abnormalities, demonstrated through high sensitivity and specificity in standalone performance and improved diagnostic accuracy in clinical studies, supporting its safety and effectiveness as a concurrent reading aid.

Sonio Suspect is a software solution designed to assist healthcare professionals in identifying and characterizing abnormal findings during or after fetal ultrasound examinations. Utilizing machine learning techniques, this web-based application automatically detects fetal ultrasound views and evaluates specific abnormal findings on uploaded images, ranging from 11 to 41 weeks of gestational age. It serves as a concurrent reading aid, presenting statuses such as "Suspected," "Not Suspected," or "Can't be analyzed" for various fetal anatomical features and abnormalities, though it emphasizes that patient management decisions should not rely solely on its analysis.A review of Sonio Suspect concluded that the device is substantially equivalent to a legally marketed predicate device, allowing it to be marketed. This determination was supported by comprehensive performance data, including software verification and validation, and extensive testing. Standalone performance testing demonstrated high sensitivity (93.2%) and specificity (90.8%) in automatically detecting abnormal fetal ultrasound findings. Furthermore, a clinical study involving multiple readers showed a significant improvement in diagnostic accuracy when readers were assisted by Sonio Suspect, reinforcing its demonstrated safety and effectiveness.

AI Regulatory Concerns

Powered by

Jev reads the full FDA letter and returns calibrated concern probabilities per regulatory category. These are AI estimates, not regulatory advice.

• Q1. What specific machine learning algorithms does Sonio Suspect employ for fetal view detection and abnormality characterization?; How does the system's performance (sensitivity, specificity) vary across different gestational ages or types of fetal abnormalities?; What are the false positive and false negative rates of Sonio Suspect, and how might these impact clinical workflow or patient anxiety?; Are there plans for Sonio Suspect to integrate with EMR systems or provide direct recommendations for further diagnostic steps?; What kind of training is recommended for clinicians to optimally utilize Sonio Suspect and integrate its findings into their diagnostic process?