Temporal Domain Guided Prediction Network
Journal
IEEE Transactions on Artificial Intelligence
Start Page
1
End Page
10
ISSN
2691-4581
Date Issued
2026
Author(s)
Abstract
Infertility affects approximately one in six reproductive-age couples globally, potentially leading to in vitro fertilization (IVF) treatment. In IVF treatment, the decision to transfer embryos on day 3 or to extend culture to day 5 remains a critical challenge for reproductive specialists. Through a guided learning approach, we propose a novel architecture named Temporal Domain Guided Prediction Network (TDGP-Net), which predicts whether the blastocyst stage quality is recommended for transfer (BRT) or not recommended for transfer (BNT) on day 3. Our experimental results demonstrate that TDGP-Net outperforms existing machine learning methods across various metrics. This advanced predictive tool can optimize embryo selection, representing a significant advancement in personalized embryo assessment.
SDGs
Publisher
Institute of Electrical and Electronics Engineers (IEEE)
Type
journal article