Skip to content
Trending storyDeveloping

Google Research tests blind sweep ultrasound AI for gestational age in new care settings

2 articles2 sourcessince Oct 21Last article 1h ago ·

Overview

AISummary of 2 articles

Google Research reports that its AI model estimating gestational age from blind sweep ultrasound kept its accuracy in two clinics, Northwestern Medicine in Chicago and Jacaranda Health in Nairobi.

The model had a mean absolute error of 4.18 days, compared with 4.48 days for expert sonographers, according to the study published in JAMA Network Open. The model was minimally tuned on only one of the target populations before being tested prospectively in both environments, with novice operators using mobile devices and unfamiliar probes.

An earlier Google Research X post said the system lets non-sonographer operators with low-cost probes produce clinical-grade gestational age assessments. That post linked to no figures or study details. The JAMA study now supplies the error figures, but the post's "clinical-grade" description is the company's own characterization.

Written by AI from the articles below · updated Oct 21, 1:16 PM ET

Check the sources:

Latest developmentOct 21, 12:26 PM ET
Google Research shows blind sweep ultrasound AI works in new care settings

Developments

2 developments

  1. Oct 21, 12:26 PM ET · 1 article
    Google Research shows blind sweep ultrasound AI generalizes to new care settings for gestational age assessment
    Google Research: Google Research shows blind sweep ultrasound AI works in new care settings
  2. Oct 21, 12:23 PM ET · 1 article
    Google Research study tests AI gestational age assessment from blind sweep ultrasound in Chicago and Nairobi clinics
    Google Research: Google Research tests AI gestational age estimation from blind sweep ultrasound across two clinics

Article timeline

The articles in this story. Times are ET.

Oct 21
  1. Google ResearchOfficial
    Google Research shows blind sweep ultrasound AI works in new care settings

    AIGoogle Research says its blind sweep ultrasound AI generalizes to new care settings. The system lets non-sonographer operators using low-cost probes produce clinical-grade gestational age assessments.

    Our blind sweep model, trained on 1,968 participants (North Carolina, US, and Zambia) was fine-tuned on scans from 120 participants from Chicago, Illinois. It was then tested prospectively in Chicago (n=192) and Nairobi (n=193), with no fine-tuning on Nairobi data. The model was non-inferior to expert sonographers at determining gestational age, in both settings, with a combined mean absolute error of 4.18 vs. 4.48 days.
  2. Google ResearchOfficialPick
    Google Research tests AI gestational age estimation from blind sweep ultrasound across two clinics

    AIGoogle Research reports that an AI model estimating gestational age from blind sweep ultrasound scans kept accuracy across Chicago and Nairobi clinics. The model reached a mean absolute error of 4.18 days, slightly better than the 4.48 days of expert sonographers, and was tested prospectively on mobile devices with novice operators and unfamiliar probes. The study was published in JAMA Network Open.

Heat Trend

Not enough continuous observations to show a trend yet.