Obstetrics / Gynecology AI medical devices

Sonio Suspect

Sonio (Samsung Medison)

CEFDA

Software that assists interpreting physicians, during or after fetal ultrasound examinations, by automatically identifying and characterising abnormal fetal ultrasound findings on detected views using machine-learning techniques. The cleared scope is eight predefined findings on six standard views of the fetal chest, abdomen and head, from 11 to 41 weeks of gestation. It is a concurrent reading aid on acquired images, not an autonomous diagnostic: each finding is flagged as suspected, not suspected or not analysable, and the clinician can override every flag. Patient management decisions should not be made solely on its results.

Recommended

Sonio Detect

Sonio (Samsung Medison)

CEFDA

Manufacturer-agnostic deep-learning software that works during a fetal ultrasound exam to automatically recognise which anatomical view is on screen, detect the structures within it, and verify the image meets the predefined quality criteria for screening. It is a concurrent reading aid — a real-time quality check on acquisition and interpretation — not an autonomous diagnostic. Cleared by the FDA (first in 2023, extended in 2024) and CE-marked Class IIa under the EU MDR in 2024; Sonio was acquired by Samsung Medison in 2024.

PeriCALM Patterns 3.0

PeriGen, Inc.

Pattern-recognition software that analyses the two channels of an electronic fetal monitoring record — the fetal heart-rate tracing and the uterine-activity recording — and annotates and summarises baseline, accelerations, decelerations and contractions as secondary adjunct information for the clinical care team during antepartum or intrapartum monitoring at 32 weeks gestation or later. One of the longest-lived computerised cardiotocography-interpretation engines on the US market: the Patterns lineage was FDA-cleared and independently benchmarked against expert panels more than a decade before the January 2025 Patterns 3.0 clearance extended its gestational scope from beyond-36 to 32-plus weeks.

Delivery Date AI

Ultrasound AI

Cloud-based post-processing software that reads standard pregnancy ultrasound images with machine learning and returns a Predicted Delivery Date. It is for singleton pregnancies between 14 0/7 and 36 6/7 weeks in women aged 18 or over who lack a reliable estimated delivery date, because the last menstrual period is unreliable or there was no first-trimester scan. The date is an aid to clinical judgment alongside standard gestational-age methods. The labelling must state that the output is not a gestational age and not a preterm-birth risk assessment.

CHLOE BLAST

Fairtility Ltd.

Machine-learning clinical decision-support software that analyses time-lapse video of cleavage-stage embryos in the IVF laboratory and predicts, on day 3, whether each embryo will develop to the blastocyst stage — an adjunct to the embryologist's own morphological grading, not a replacement for it. CHLOE BLAST is the module of Fairtility's CHLOE platform cleared by the FDA in August 2025, for use only with images from the EmbryoScope version D incubator; the broader CHLOE EQ decision-support module has carried a CE mark under the EU MDR since 2022 and is in commercial use across Europe and the Asia-Pacific region.

Butterfly Gestational Age Tool

Butterfly Network, Inc.

Software feature inside the Butterfly iQ app that estimates gestational age from blind ultrasound sweeps of the pregnant abdomen. The operator measures fundal height, which sets how many sweeps are needed, then moves an iQ+ or iQ3 probe along animated paths without reading the live image. A deep-learning model trained at the University of North Carolina turns the sweep cines into one gestational-age estimate. No fetal biometry is measured. The clearance covers singleton intrauterine pregnancies presumed to be between 16 and 37 weeks, as adjunctive information for trained healthcare professionals, and not for prenatal management or delivery planning.