Real-world use of an artificial intelligence–powered clinical decision support tool for ovarian stimulation

Objective: To understand how treatment decisions and patient outcomes change with physician utilization of artificial intelligence (AI) to help determine follicle-stimulating hormone (FSH) starting dose and trigger injection timing during ovarian stimulation. Design: Retrospective cohort study with...

Full description

Saved in:
Bibliographic Details
Main Authors: Cameron J. Bixby, D.O., Bradley Miller, M.D.
Format: Article
Language:English
Published: Elsevier 2025-06-01
Series:F&S Reports
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2666334125000170
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849689225644998656
author Cameron J. Bixby, D.O.
Bradley Miller, M.D.
author_facet Cameron J. Bixby, D.O.
Bradley Miller, M.D.
author_sort Cameron J. Bixby, D.O.
collection DOAJ
description Objective: To understand how treatment decisions and patient outcomes change with physician utilization of artificial intelligence (AI) to help determine follicle-stimulating hormone (FSH) starting dose and trigger injection timing during ovarian stimulation. Design: Retrospective cohort study with historical controls. Subjects: Patients undergoing ovarian stimulation by multiple physicians at one in vitro fertilization clinic in the United States. Exposure: A total of 292 patients were treated between December 2022 and December 2023 with adjunctive clinician use of AI to help select starting dose of FSH and timing of the trigger injection. These were matched to 292 historical control patients treated between May 2019 and May 2022 by the same physicians without AI. Main Outcome Measures: The primary endpoints were the starting FSH dose, total FSH dose, and number of metaphase II (MII) oocytes retrieved at the end of stimulation. Results: The use of AI did not introduce any adverse events. After patient matching, there were no statistically significant differences in age, body mass index, antimüllerian hormone, or antral follicle count between the treatment and control groups. Comparing the treatment arm with the control arm, the average number of MII oocytes was 11.17 vs. 11.25, the average starting FSH dose was 397.09 IU vs. 443.84 IU, and the average total FSH dose was 4,181.77 IU vs. 4,654.71 IU. Conclusion: Physician use of AI helped significantly reduce the starting and total FSH doses prescribed to patients without adversely affecting MII outcomes, indicating the potential use of AI in lowering in vitro fertilization costs to patients.
format Article
id doaj-art-8af0a4a26e5c48ae9ecd27e1a18cd693
institution DOAJ
issn 2666-3341
language English
publishDate 2025-06-01
publisher Elsevier
record_format Article
series F&S Reports
spelling doaj-art-8af0a4a26e5c48ae9ecd27e1a18cd6932025-08-20T03:21:42ZengElsevierF&S Reports2666-33412025-06-016214014610.1016/j.xfre.2025.01.015Real-world use of an artificial intelligence–powered clinical decision support tool for ovarian stimulationCameron J. Bixby, D.O.0Bradley Miller, M.D.1Henry Ford Providence Hospital, Michigan State University College of Human Medicine, Southfield, Michigan; Correspondence: Cameron J. Bixby, D.O., Henry Ford Providence Hospital, Michigan State University College of Human Medicine, Southfield, Michigan 48075.Reproductive Medicine Associates of Michigan, Reproductive Endocrinology, Troy, MichiganObjective: To understand how treatment decisions and patient outcomes change with physician utilization of artificial intelligence (AI) to help determine follicle-stimulating hormone (FSH) starting dose and trigger injection timing during ovarian stimulation. Design: Retrospective cohort study with historical controls. Subjects: Patients undergoing ovarian stimulation by multiple physicians at one in vitro fertilization clinic in the United States. Exposure: A total of 292 patients were treated between December 2022 and December 2023 with adjunctive clinician use of AI to help select starting dose of FSH and timing of the trigger injection. These were matched to 292 historical control patients treated between May 2019 and May 2022 by the same physicians without AI. Main Outcome Measures: The primary endpoints were the starting FSH dose, total FSH dose, and number of metaphase II (MII) oocytes retrieved at the end of stimulation. Results: The use of AI did not introduce any adverse events. After patient matching, there were no statistically significant differences in age, body mass index, antimüllerian hormone, or antral follicle count between the treatment and control groups. Comparing the treatment arm with the control arm, the average number of MII oocytes was 11.17 vs. 11.25, the average starting FSH dose was 397.09 IU vs. 443.84 IU, and the average total FSH dose was 4,181.77 IU vs. 4,654.71 IU. Conclusion: Physician use of AI helped significantly reduce the starting and total FSH doses prescribed to patients without adversely affecting MII outcomes, indicating the potential use of AI in lowering in vitro fertilization costs to patients.http://www.sciencedirect.com/science/article/pii/S2666334125000170Artificial intelligencemachine learningovarian stimulationART
spellingShingle Cameron J. Bixby, D.O.
Bradley Miller, M.D.
Real-world use of an artificial intelligence–powered clinical decision support tool for ovarian stimulation
F&S Reports
Artificial intelligence
machine learning
ovarian stimulation
ART
title Real-world use of an artificial intelligence–powered clinical decision support tool for ovarian stimulation
title_full Real-world use of an artificial intelligence–powered clinical decision support tool for ovarian stimulation
title_fullStr Real-world use of an artificial intelligence–powered clinical decision support tool for ovarian stimulation
title_full_unstemmed Real-world use of an artificial intelligence–powered clinical decision support tool for ovarian stimulation
title_short Real-world use of an artificial intelligence–powered clinical decision support tool for ovarian stimulation
title_sort real world use of an artificial intelligence powered clinical decision support tool for ovarian stimulation
topic Artificial intelligence
machine learning
ovarian stimulation
ART
url http://www.sciencedirect.com/science/article/pii/S2666334125000170
work_keys_str_mv AT cameronjbixbydo realworlduseofanartificialintelligencepoweredclinicaldecisionsupporttoolforovarianstimulation
AT bradleymillermd realworlduseofanartificialintelligencepoweredclinicaldecisionsupporttoolforovarianstimulation