Leveraging Attention Mechanisms for Interpretable Human Embryo Image Segmentation

Wided Miled, Wided Miled, Nozha Chakroun

2025

Abstract

In-vitro Fertilization (IVF) is a widely used assisted reproductive technology where embryos are cultured under controlled laboratory conditions. The selection of a high-quality blastocyst, typically reached five days after fertilization, is crucial to the success of the IVF procedure. Therefore, evaluating embryo quality at this stage is essential to optimize IVF outcomes. Advances in neural network architectures, particularly Convolutional Neural Networks (CNNs), have enhanced decision-making in IVF. However, ensuring both accuracy and interpretability in these models remains a challenge. This paper focuses on improving human blastocyst segmentation by combining channel attention mechanisms with a ResNet50 model within an encoder-decoder architecture. The method accurately identifies key blastocyst components such as inner cell mass (ICM), trophectoderm (TE), and zona pellucida (ZP). Our approach was validated on a publicly available human embryo dataset, achieving Intersection over Union (IoU) scores of 83.09% for ICM, 86.87% for ZP, and 81.1% for TE, outperforming current state-of-the-art methods. These results demonstrate the potential of deep learning to improve both accuracy and interpretability in embryo quality assessment.

Paper Citation

in Harvard Style

Miled W. and Chakroun N. (2025). Leveraging Attention Mechanisms for Interpretable Human Embryo Image Segmentation. In Proceedings of the 17th International Conference on Agents and Artificial Intelligence - Volume 2: ICAART; ISBN 978-989-758-737-5, SciTePress, pages 872-879. DOI: 10.5220/0013383200003890

in Bibtex Style

@conference{icaart25,
author={Wided Miled and Nozha Chakroun},
title={Leveraging Attention Mechanisms for Interpretable Human Embryo Image Segmentation},
booktitle={Proceedings of the 17th International Conference on Agents and Artificial Intelligence - Volume 2: ICAART},
year={2025},
pages={872-879},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013383200003890},
isbn={978-989-758-737-5},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 17th International Conference on Agents and Artificial Intelligence - Volume 2: ICAART
TI - Leveraging Attention Mechanisms for Interpretable Human Embryo Image Segmentation
SN - 978-989-758-737-5
AU - Miled W.
AU - Chakroun N.
PY - 2025
SP - 872
EP - 879
DO - 10.5220/0013383200003890
PB - SciTePress