Modelling follicular growth during ovarian stimulation using agent-based artificial intelligence

Artsiom Hramyka; Thomas W Kelsey; Simon Hanassab; Scott M Nelson; Arthur C. Yeung; Sotirios Saravelos; Rehan Salim; Alexander N. Comninos; Krasimira Tsaneva-Atanasova; Margaritis Voliotis; Geoffrey H. Trew; Thomas Heinis; Waljit S. Dhillo; Ali Abbara

doi:10.1210/clinem/dgaf539

Modelling follicular growth during ovarian stimulation using agent-based artificial intelligence

Artsiom Hramyka, Thomas W Kelsey^*, Simon Hanassab, Scott M Nelson, Arthur C. Yeung, Sotirios Saravelos, Rehan Salim, Alexander N. Comninos, Krasimira Tsaneva-Atanasova, Margaritis Voliotis, Geoffrey H. Trew, Thomas Heinis, Waljit S. Dhillo, Ali Abbara^*

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

Abstract

Context: Ovarian stimulation is a key step in medically assisted reproduction (MAR), whereby supraphysiological doses of FSH extend the “FSH window” and induce multifollicular growth. However, only limited data exist that examine individual follicular growth rates during fertility treatment.

Objective: To model growth rates of individual ovarian follicles during ovarian stimulation in MAR cycles using an agent-based artificial intelligence model.

Design: Observational cohort study.

Setting: Eleven assisted conception clinics in Europe.

Patients: 11 572 patients (2005-2023) who underwent ovarian stimulation during MAR.

Intervention: Predictive modeling was conducted using 39 698 scans including 434 082 follicles from 12 950 cycles during ovarian stimulation.

Main Outcome Measures: Daily growth rates of individual ovarian follicles during stimulation were modeled to enable prediction of follicle sizes at the end of ovarian stimulation.

Results: Mean follicle growth rate of ovarian follicles was 1.350 mm/day (95% CI: 1.346–1.353 mm/day) and was significantly associated with antral follicle count and FSH dose changes (both P < .001). Using only the first scan, the model enabled prediction of follicles sizes within 2 mm at the end of ovarian stimulation with 75.0% accuracy (95% CI: 74.6–75.3%), increasing to 80.1% (95% CI: 79.8–80.5%) when incorporating the first 2 scans. Predictive performance was stable across clinics, with a mean accuracy of 78.0% in a random training-test split, and 77.1% using cross-validation by clinic.

Conclusion: We used advanced artificial intelligence techniques to progress our understanding of follicle growth dynamics during ovarian stimulation. This model can reliably predict follicle size profiles at the end of stimulation enabling moderation of the number of scans required.

Original language	English
Pages (from-to)	615-621
Number of pages	7
Journal	Journal of Clinical Endocrinology & Metabolism
Volume	111
Issue number	3
Early online date	29 Oct 2025
DOIs	https://doi.org/10.1210/clinem/dgaf539
Publication status	Published - 1 Mar 2026

Keywords

Follicle growth
Reproductive endocrinology
Ovarian stimulation
Artificial intelligence
Machine learning
Ultrasound

Access to Document

10.1210/clinem/dgaf539Licence: CC BY

Hramyka_2026_JCEM_Modelling-follicular-growth-ovarian-stimulation_CC
© The Author(s) 2025. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.
Final published version, 632 KBLicence: CC BY

Cite this

Hramyka, A., Kelsey, T. W., Hanassab, S., Nelson, S. M., Yeung, A. C., Saravelos, S., Salim, R., Comninos, A. N., Tsaneva-Atanasova, K., Voliotis, M., Trew, G. H., Heinis, T., Dhillo, W. S., & Abbara, A. (2026). Modelling follicular growth during ovarian stimulation using agent-based artificial intelligence. Journal of Clinical Endocrinology & Metabolism, 111(3), 615-621. https://doi.org/10.1210/clinem/dgaf539

@article{d6d1f7affb204a06bdfc62a13d49da0d,

title = "Modelling follicular growth during ovarian stimulation using agent-based artificial intelligence",

abstract = "Context: Ovarian stimulation is a key step in medically assisted reproduction (MAR), whereby supraphysiological doses of FSH extend the “FSH window” and induce multifollicular growth. However, only limited data exist that examine individual follicular growth rates during fertility treatment. Objective: To model growth rates of individual ovarian follicles during ovarian stimulation in MAR cycles using an agent-based artificial intelligence model. Design: Observational cohort study. Setting: Eleven assisted conception clinics in Europe. Patients: 11 572 patients (2005-2023) who underwent ovarian stimulation during MAR. Intervention: Predictive modeling was conducted using 39 698 scans including 434 082 follicles from 12 950 cycles during ovarian stimulation. Main Outcome Measures: Daily growth rates of individual ovarian follicles during stimulation were modeled to enable prediction of follicle sizes at the end of ovarian stimulation. Results: Mean follicle growth rate of ovarian follicles was 1.350 mm/day (95\% CI: 1.346–1.353 mm/day) and was significantly associated with antral follicle count and FSH dose changes (both P < .001). Using only the first scan, the model enabled prediction of follicles sizes within 2 mm at the end of ovarian stimulation with 75.0\% accuracy (95\% CI: 74.6–75.3\%), increasing to 80.1\% (95\% CI: 79.8–80.5\%) when incorporating the first 2 scans. Predictive performance was stable across clinics, with a mean accuracy of 78.0\% in a random training-test split, and 77.1\% using cross-validation by clinic. Conclusion: We used advanced artificial intelligence techniques to progress our understanding of follicle growth dynamics during ovarian stimulation. This model can reliably predict follicle size profiles at the end of stimulation enabling moderation of the number of scans required.",

keywords = "Follicle growth, Reproductive endocrinology, Ovarian stimulation, Artificial intelligence, Machine learning, Ultrasound",

author = "Artsiom Hramyka and Kelsey, \{Thomas W\} and Simon Hanassab and Nelson, \{Scott M\} and Yeung, \{Arthur C.\} and Sotirios Saravelos and Rehan Salim and Comninos, \{Alexander N.\} and Krasimira Tsaneva-Atanasova and Margaritis Voliotis and Trew, \{Geoffrey H.\} and Thomas Heinis and Dhillo, \{Waljit S.\} and Ali Abbara",

note = "Funding: The Section of Endocrinology and Investigative Medicine (Imperial College London) is funded by grants from the MRC and NIHR and is supported by the NIHR Biomedical Research Centre Funding Scheme and the NIHR Imperial Clinical Research Facility. S.H. is supported by the UKRI CDT in AI for Healthcare http://ai4health.io (EP/S023283/1). W.S.D. is supported by an NIHR Senior Investigator Award (NIHR202371). A.A. is supported by an NIHR Clinician Scientist Award (CS-2018-18-ST2-002). ",

year = "2026",

month = mar,

day = "1",

doi = "10.1210/clinem/dgaf539",

language = "English",

volume = "111",

pages = "615--621",

journal = "Journal of Clinical Endocrinology \& Metabolism",

issn = "0021-972X",

publisher = "Endocrine Society",

number = "3",

}

Hramyka, A, Kelsey, TW, Hanassab, S, Nelson, SM, Yeung, AC, Saravelos, S, Salim, R, Comninos, AN, Tsaneva-Atanasova, K, Voliotis, M, Trew, GH, Heinis, T, Dhillo, WS & Abbara, A 2026, 'Modelling follicular growth during ovarian stimulation using agent-based artificial intelligence', Journal of Clinical Endocrinology & Metabolism, vol. 111, no. 3, pp. 615-621. https://doi.org/10.1210/clinem/dgaf539

TY - JOUR

T1 - Modelling follicular growth during ovarian stimulation using agent-based artificial intelligence

AU - Hramyka, Artsiom

AU - Kelsey, Thomas W

AU - Hanassab, Simon

AU - Nelson, Scott M

AU - Yeung, Arthur C.

AU - Saravelos, Sotirios

AU - Salim, Rehan

AU - Comninos, Alexander N.

AU - Tsaneva-Atanasova, Krasimira

AU - Voliotis, Margaritis

AU - Trew, Geoffrey H.

AU - Heinis, Thomas

AU - Dhillo, Waljit S.

AU - Abbara, Ali

N1 - Funding: The Section of Endocrinology and Investigative Medicine (Imperial College London) is funded by grants from the MRC and NIHR and is supported by the NIHR Biomedical Research Centre Funding Scheme and the NIHR Imperial Clinical Research Facility. S.H. is supported by the UKRI CDT in AI for Healthcare http://ai4health.io (EP/S023283/1). W.S.D. is supported by an NIHR Senior Investigator Award (NIHR202371). A.A. is supported by an NIHR Clinician Scientist Award (CS-2018-18-ST2-002).

PY - 2026/3/1

Y1 - 2026/3/1

N2 - Context: Ovarian stimulation is a key step in medically assisted reproduction (MAR), whereby supraphysiological doses of FSH extend the “FSH window” and induce multifollicular growth. However, only limited data exist that examine individual follicular growth rates during fertility treatment. Objective: To model growth rates of individual ovarian follicles during ovarian stimulation in MAR cycles using an agent-based artificial intelligence model. Design: Observational cohort study. Setting: Eleven assisted conception clinics in Europe. Patients: 11 572 patients (2005-2023) who underwent ovarian stimulation during MAR. Intervention: Predictive modeling was conducted using 39 698 scans including 434 082 follicles from 12 950 cycles during ovarian stimulation. Main Outcome Measures: Daily growth rates of individual ovarian follicles during stimulation were modeled to enable prediction of follicle sizes at the end of ovarian stimulation. Results: Mean follicle growth rate of ovarian follicles was 1.350 mm/day (95% CI: 1.346–1.353 mm/day) and was significantly associated with antral follicle count and FSH dose changes (both P < .001). Using only the first scan, the model enabled prediction of follicles sizes within 2 mm at the end of ovarian stimulation with 75.0% accuracy (95% CI: 74.6–75.3%), increasing to 80.1% (95% CI: 79.8–80.5%) when incorporating the first 2 scans. Predictive performance was stable across clinics, with a mean accuracy of 78.0% in a random training-test split, and 77.1% using cross-validation by clinic. Conclusion: We used advanced artificial intelligence techniques to progress our understanding of follicle growth dynamics during ovarian stimulation. This model can reliably predict follicle size profiles at the end of stimulation enabling moderation of the number of scans required.

AB - Context: Ovarian stimulation is a key step in medically assisted reproduction (MAR), whereby supraphysiological doses of FSH extend the “FSH window” and induce multifollicular growth. However, only limited data exist that examine individual follicular growth rates during fertility treatment. Objective: To model growth rates of individual ovarian follicles during ovarian stimulation in MAR cycles using an agent-based artificial intelligence model. Design: Observational cohort study. Setting: Eleven assisted conception clinics in Europe. Patients: 11 572 patients (2005-2023) who underwent ovarian stimulation during MAR. Intervention: Predictive modeling was conducted using 39 698 scans including 434 082 follicles from 12 950 cycles during ovarian stimulation. Main Outcome Measures: Daily growth rates of individual ovarian follicles during stimulation were modeled to enable prediction of follicle sizes at the end of ovarian stimulation. Results: Mean follicle growth rate of ovarian follicles was 1.350 mm/day (95% CI: 1.346–1.353 mm/day) and was significantly associated with antral follicle count and FSH dose changes (both P < .001). Using only the first scan, the model enabled prediction of follicles sizes within 2 mm at the end of ovarian stimulation with 75.0% accuracy (95% CI: 74.6–75.3%), increasing to 80.1% (95% CI: 79.8–80.5%) when incorporating the first 2 scans. Predictive performance was stable across clinics, with a mean accuracy of 78.0% in a random training-test split, and 77.1% using cross-validation by clinic. Conclusion: We used advanced artificial intelligence techniques to progress our understanding of follicle growth dynamics during ovarian stimulation. This model can reliably predict follicle size profiles at the end of stimulation enabling moderation of the number of scans required.

KW - Follicle growth

KW - Reproductive endocrinology

KW - Ovarian stimulation

KW - Artificial intelligence

KW - Machine learning

KW - Ultrasound

U2 - 10.1210/clinem/dgaf539

DO - 10.1210/clinem/dgaf539

M3 - Article

SN - 0021-972X

VL - 111

SP - 615

EP - 621

JO - Journal of Clinical Endocrinology & Metabolism

JF - Journal of Clinical Endocrinology & Metabolism

IS - 3

ER -

Modelling follicular growth during ovarian stimulation using agent-based artificial intelligence

Abstract

Keywords

Access to Document

Fingerprint

Cite this