This white paper presents how ViewAssist™ and BiometryAssist™, AI-powered tools built into Samsung Medison’s HERA series systems, support more accurate and efficient fetal brain ultrasound in clinical practice. These technologies automate the recognition, annotation, and measurement of fetal CNS structures—minimizing operator dependency and reducing scan time in obstetric ultrasound workflows.
The system automatically identifies standard axial brain planes—transventricular (TV), transthalamic (TT), and transcerebellar (TC)—and applies measurements such as lateral ventricle width, biparietal diameter, head circumference, and nuchal fold thickness. All annotations and measurements follow ISUOG guidelines to ensure consistency and accuracy.
Validated using 890 fetal ultrasound images, ViewAssist™ and BiometryAssist™ achieved a 94.5% plane recognition rate and over 98% accuracy in automated caliper placement. Their implementation reduced keystrokes and measurement time significantly, offering clinical efficiency and standardization.
The tools performed particularly well in second-trimester scans, indicating the potential for further enhancement through expanded training data. Future development may also include built-in image validation to boost reliability.
ViewAssist™ and BiometryAssist™ are efficient and intelligent tools for fetal CNS evaluation that can improve work efficiency and reduce operator dependency.
For practical steps on how to apply these features, visit the Quick Guide to ViewAssist™ .