Search

Infertility affects 1-in-6 couples, with repeated intensive cycles of assisted reproductive technology (ART) required by many to achieve a desired live birth. In ART, typically, clinicians and laboratory staff consider patient characteristics, previous treatment responses, and ongoing monitoring to determine treatment decisions. However, the reproducibility, weighting, and interpretation of these characteristics are contentious, and highly operator-dependent, resulting in considerable reliance on clinical experience. Artificial intelligence (AI) is ideally suited to handle, process, and analyze large, dynamic, temporal datasets with multiple intermediary outcomes that are generated during an ART cycle. Here, we review how AI has demonstrated potential for optimization and personalization of key steps in a reproducible manner, including: drug selection and dosing, cycle monitoring, induction of oocyte maturation, and selection of the most competent gametes and embryos, to improve the overall efficacy and safety of ART., (© 2024. The Author(s).)

Objective: To quantify how representative a single measure of reproductive hormone level is of the daily hormonal profile using data from detailed hormonal sampling in the saline placebo-treated arm conducted over several hours., Design: Retrospective analysis of data from previous interventional research studies evaluating reproductive hormones., Setting: Clinical Research Facility at a tertiary reproductive endocrinology centre at Imperial College Hospital NHS Foundation Trust., Patients: Overall, 266 individuals, including healthy men and women (n = 142) and those with reproductive disorders and states (n = 124 [11 with functional hypothalamic amenorrhoea, 6 with polycystic ovary syndrome, 62 women and 32 men with hypoactive sexual desire disorder, and 13 postmenopausal women]), were included in the analysis., Interventions: Data from 266 individuals who had undergone detailed hormonal sampling in the saline placebo-treated arms of previous research studies was used to quantify the variability in reproductive hormones because of pulsatile secretion, diurnal variation, and feeding using coefficient of variation (CV) and entropy., Main Outcome Measures: The ability of a single measure of reproductive hormone level to quantify the variability in reproductive hormone levels because of pulsatile secretion, diurnal variation, and nutrient intake., Results: The initial morning value of reproductive hormone levels was typically higher than the mean value throughout the day (percentage decrease from initial morning measure to daily mean: luteinizing hormone level 18.4%, follicle-stimulating hormone level 9.7%, testosterone level 9.2%, and estradiol level 2.1%). Luteinizing hormone level was the most variable (CV 28%), followed by sex-steroid hormone levels (testosterone level 12% and estradiol level 13%), whereas follicle-stimulating hormone level was the least variable reproductive hormone (CV 8%). In healthy men, testosterone levels fell between 9:00 am and 5:00 pm by 14.9% (95% confidence interval 4.2, 25.5%), although morning levels correlated with (and could be predicted from) late afternoon levels in the same individual (r 2 = 0.53, P<.0001). Testosterone levels were reduced more after a mixed meal (by 34.3%) than during ad libitum feeding (9.5%), after an oral glucose load (6.0%), or an intravenous glucose load (7.4%)., Conclusion: Quantification of the variability of a single measure of reproductive hormone levels informs the reliability of reproductive hormone assessment., Competing Interests: Declaration of interests A.A. was supported by NIHR Clinician Scientist Award CS-2018-18-ST2-002. S.A. has nothing to disclose. M.P. was supported by an NIHR Academic Clinical Lectureship Award. C.I.E. was supported by an Imperial-BRC IPPRF Award (P79696). E.G.M. was supported by an NIHR Academic Clinical Lectureship Award. L.T. has nothing to disclose. K.K. was supported by NIHR Academic Clinical Fellowship Award ACF-2021-21-001. S.H. was supported by the UKRI CDT in AI for Healthcare http://ai4health.io (Grant number EP/S023283/1). T.H. has nothing to disclose. T.M.-M.T. was supported by the NIHR, Diabetes UK, and the JP Moulton Charitable Trust. K.T.A. was supported by the EPSRC via grant EP/T017856/1. A.N.C. was supported by the NHS. M.V. has nothing to disclose. W.S.D. was supported by an NIHR Senior Investigator Award., (Copyright © 2024. Published by Elsevier Inc.)