EUVIMEDEUVIMED entdecken
Crossref

L26/P-862 Environmental Determinants of IVF Success: Evidence from Laboratory Data

R. Khoshkangini, Z. Asghari Varzaneh, T. Ebner

Human Reproduction · 01.07.2026 · 41 (Supplement_1)

DOI: 10.1093/humrep/deag083.1187 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Human Reproduction, online erschienen am 08.07.2026. Verfasst von R. Khoshkangini und 2 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Study question To what extent do environmental factors—including humidity, temperature, VOCs, etc. affect in vitro fertilization (IVF) success rates, as revealed through explainable artificial intelligence (XAI) methods? Summary answer Environmental factors influence IVF success differently across maternal age groups, with distinct patterns of impact observed for specific environmental variables. What is known already Previous studies suggest that environmental conditions within IVF laboratories may influence reproductive outcomes. However, existing research typically examines individual factors in isolation and lacks quantitative assessments of their combined and time-dependent effects. Moreover, there is a limited understanding of how these environmental factors interact with patient characteristics, such as age. No prior study has comprehensively quantified multiple environmental determinants simultaneously using temporal machine learning models combined with explainable AI techniques. Study design, size, duration The three-year study is organized into multiple work packages and is currently in its mid-project phase. The ongoing work focuses on ensemble-based predictive modeling and result interpretation using explainable AI (XAI). In the forthcoming phase, embryo image data will be integrated as an additional modality to uncover patterns associated with IVF success rates. Participants/materials, setting, methods This study was conducted as part of an international collaboration involving six partner organizations from academia and IVF clinics. Environmental sensor data and clinical outcome data were continuously collected from IVF laboratories located outside Europe. The data were analyzed using ensemble-based machine learning models, augmented with explainable AI (XAI) techniques to explain the results. In the next phase multimodal approach will be employed to integrate information from multiple sources to uncover complex patterns. Main results and the role of chance Environmental and clinical data were collected from around 1,000 IVF patients in 2025 across several clinics located in relatively cold-climate regions (latitude ≈ 34.31°, longitude ≈ 47.07°, elevation ≈1,350). The dataset combined detailed patient clinical characteristics with high-resolution environmental measurements obtained directly from IVF laboratory environments. In line with established clinical evidence, maternal age was confirmed as the most influential predictor of pregnancy success, reinforcing its central role in IVF outcomes. However, beyond age, multiple environmental factors showed statistically significant and clinically relevant associations with treatment success. To analyze these effects, an ensemble decision-based ML model integrating clinical, environmental, and temporal variables was developed. The model demonstrated strong predictive performance (R² = 0.95, RMSE = 0.096), indicating high accuracy in capturing outcome variability. XAI analysis identified temporal features and ambient temperature as key contributors, emphasizing the relevance of seasonal and time-dependent effects. When seasonal indicators were removed, laboratory environmental conditions, including temperature, relative humidity, and VOC, became the dominant predictors. Notably, their importance differed across maternal age groups, suggesting age-dependent sensitivity to environmental factors. Although these findings are promising, data were collected within a limited climatic range, and further studies across diverse regions are required to confirm robustness and generalizability. Limitations, reasons for caution The primary limitation of this study is the availability and heterogeneity of data across clinics. Although efforts were made to include additional centers and install more sensors, broader deployment is required to ensure greater generalizability of the findings. Wider implications of the findings This study provides quantitative evidence that IVF laboratory environmental conditions influence treatment success in an age-dependent manner. The results highlight the potential of XAI-driven decision support systems to optimize laboratory environments and improve IVF outcomes. Trial registration number No

Zitieren

APA

Khoshkangini, R., Varzaneh, Z. A., & Ebner, T. (2026). L26/P-862 Environmental Determinants of IVF Success: Evidence from Laboratory Data. Human Reproduction, 41(Supplement_1). https://doi.org/10.1093/humrep/deag083.1187

Vancouver

Khoshkangini R, Varzaneh ZA, Ebner T. L26/P-862 Environmental Determinants of IVF Success: Evidence from Laboratory Data. Human Reproduction. 2026;41(Supplement_1). doi:10.1093/humrep/deag083.1187

BibTeX
@article{Khoshkangini2026,
  author = {Khoshkangini, R. and Varzaneh, Z. Asghari and Ebner, T.},
  title = {L26/P-862 Environmental Determinants of IVF Success: Evidence from Laboratory Data},
  journal = {Human Reproduction},
  year = {2026},
  volume = {41},
  number = {Supplement_1},
  doi = {10.1093/humrep/deag083.1187}
}
Kostenlos starten →