Skip to main content
SMFM 2026 Pregnancy Meeting Main banner
Icon Legend
This session is not in your schedule.
This session is in your schedule. Click again to remove it.
Presentation Icons
Additional registration fee required
Additional registration fee required
Faculty have requested this content not be shared outside of the session
Faculty have requested this content not be shared outside of the session
CME Credit Offered
CME Credit Offered
Abstract Award
Abstract Award
Recording available 2/16-5/2
Recording available 2/16-5/2
AIUM Credit
AIUM Credit
Foundation Awardee
Foundation Awardee
Poster Icons
Abstract Award
Abstract Award
Foundation Awardee
Foundation Awardee

Poster Session 2

Category: Digital Health Technologies (DHT)

Poster Session 2

(529) How does ChatGPT compare to MFM specialists in diagnosing PAS?

Wednesday, February 11, 2026
3:30 PM - 5:00 PM  

    Submitting Author and Presenting Author(s)

  • Joe Haydamous, MD (he/him/his)

    PGY1
    Department of Obstetrics, Gynecology and Reproductive Sciences, McGovern Medical School at UTHealth Houston
    Department of Obstetrics and Gynecology, McGovern Medical School at UT Health, Houston, Texas, United States

    • Coauthor(s)

  • Edgar A. Hernandez-Andrade, MD, PhD (he/him/his)

    Professor
    Division of Maternal-Fetal Medicine, Department of Obstetrics, Gynecology and Reproductive Sciences, McGovern Medical School at UTHealth Houston
    Houston, Texas, United States

  • Laura Diab, MD (she/her/hers)

    Division of Maternal-Fetal Medicine, Department of Obstetrics, Gynecology and Reproductive Sciences, McGovern Medical School at UTHealth Houston
    Houston, Texas, United States

  • Baha M. Sibai, MD

    Professor
    Division of Maternal-Fetal Medicine, Department of Obstetrics, Gynecology and Reproductive Sciences, McGovern Medical School at UTHealth Houston
    Houston, Texas, United States

  • Farah H. Amro, MD

    Assistant Professor
    Division of Maternal-Fetal Medicine, Department of Obstetrics, Gynecology and Reproductive Sciences, McGovern Medical School at UTHealth Houston
    Houston, Texas, United States

Objective:
ChatGPT (Generative Pre‑trained Transformer) is a deep‑learning technology with potential healthcare applications. We evaluated its diagnostic performance for placenta accreta spectrum (PAS) on ultrasound (US) versus maternal‑fetal medicine (MFM) specialists, using pathology as the gold standard, and examined performance across transabdominal (TA), transvaginal (TV), and combined (TA+TV) imaging.

Study Design:
This retrospective study included patients referred for suspected PAS who underwent US at our institution. All non‑PAS cases had placenta previa; PAS cases included previa and non‑previa. US images and video clips were reviewed by MFM specialists and ChatGPT using standardized prompts, blinded to outcome. Interpretations were classified as PAS present (yes/no) and subtype. Pathology was the reference standard. Agreement was assessed with Cohen’s kappa (κ). Sensitivity, specificity, PPV, NPV, and accuracy were calculated. McNemar’s test compared modalities.

Results:
Thirty cases were evaluated; 13 (43.3%) were non‑PAS. Among PAS cases (n=17), MFMs showed substantial agreement with pathology (κ=0.798,p< 0.001); ChatGPT showed moderate agreement (κ=0.609,p< 0.001). MFMs and ChatGPT had substantial mutual agreement (κ=0.627,p< 0.001). MFMs sensitivity and accuracy exceeded ChatGPT (88.2% vs. 68.6% and 90.0% vs. 80.0%), though specificity was high for both ( >92%). ChatGPT correctly identified 12/13 non‑PAS placenta previa cases (specificity 92.3%) with an NPV of 80.0%, supporting its potential as a rule‑out tool. Sensitivity improved from 47.1% with TA alone to 76.5% with TV and 82.4% with TA+TV, with accuracy rising from 70.0% to 83.3% and 86.7%. No modality differences were statistically significant (all p >0.05) but trends favored TA+TV and TV.

Conclusion:

While MFM specialists outperformed ChatGPT in sensitivity and accuracy, ChatGPT’s high specificity and NPV support exploring its possible role to aid in the community setting. While not statistically significant, performance overall improved with the addition of transvaginal imaging, underscoring its value in evaluating the PAS patient.