BELA

Accurate and Noninvasive Ploidy Prediction for Human Preimplantation Embryos

Project Description

This project aims to enhance the assessment of fertilized human embryos, a critical step in the process of in vitro fertilization (VF). The research paper behind this project introduced BELA, the Blastocyst Evaluation Learning Algorithm, a novel model for embryo ploidy status prediction. It surpasses both image- and video-based ploidy models without necessitating any subjective input from embryologists. This project uses deep learning techniques to accurately and noninvasively predict ploidy in preimplantation human embryos.

Contributors

• Suraj Rajendran, Institute for Computational Biomedicine, Department of Physiology and Biophysics, Weill Cornell Medicine of Cornell University, New York, NY, USA, Tri-Institutional Computational Biology & Medicine Program, Cornell University, NY, USA
• Matthew Brendel, Institute for Computational Biomedicine, Department of Physiology and Biophysics, Weill Cornell Medicine of Cornell University, New York, NY, USA
• Josue Barnes, Institute for Computational Biomedicine, Department of Physiology and Biophysics, Weill Cornell Medicine of Cornell University, New York, NY, USA
• Qiansheng Zhan, The Ronald O. Perelman and Claudia Cohen Center for Reproductive Medicine, Weill Cornell Medicine, New York, NY, USA
• Jonas E. Malmsten, The Ronald O. Perelman and Claudia Cohen Center for Reproductive Medicine, Weill Cornell Medicine, New York, NY, USA
• Pantelis Zisimopoulos, Institute for Computational Biomedicine, Department of Physiology and Biophysics, Weill Cornell Medicine of Cornell University, New York, NY, USA
• Alexandros Sigaras, Institute for Computational Biomedicine, Department of Physiology and Biophysics, Weill Cornell Medicine of Cornell University, New York, NY, USA
• Marcos Meseguer, IVI Valencia, Health Research Institute la Fe, Valencia, Spain
• Kathleen A Miller, IVF Florida Reproductive Associates, Fort Lauderdale, Florida, USA
• David Hoffman, IVF Florida Reproductive Associates, Fort Lauderdale, Florida, USA
• Zev Rosenwaks, The Ronald O. Perelman and Claudia Cohen Center for Reproductive Medicine, Weill Cornell Medicine, New York, NY, USA
• Olivier Elemento, Institute for Computational Biomedicine, Department of Physiology and Biophysics, Weill Cornell Medicine of Cornell University, New York, NY, USA
• Nikica Zaninovic, The Ronald O. Perelman and Claudia Cohen Center for Reproductive Medicine, Weill Cornell Medicine, New York, NY, USA
• Iman Hajirasouliha (Corresponding author), Institute for Computational Biomedicine, Department of Physiology and Biophysics, Weill Cornell Medicine of Cornell University, New York, NY, USA, imh2003@med.cornell.edu

Code Description

The code in this repository forms the backbone of the BELA model. It is designed to use time-lapse imaging to predict embryo ploidy status and identify critical time points during embryo development that would maximize prediction accuracy. Specifically, it leverages deep learning techniques and uses video classification models to outperform single image classification models.

Instructions for Use

The BELA system uses several Python packages. Before running any of the scripts, ensure the following dependencies are installed in your Python environment.

Files available provide code for training and evaluating a BELA model through:

1. Video Creation
2. Annotation File Creation
3. Training and Prediction

Contact

For any questions or comments, please contact Iman Hajirasouliha at imh2003@med.cornell.edu.

Citation

If you use this code for your research, please cite:

Rajendran, S., Brendel, M., Barnes, J., Zhan, Q., Malmsten, J.E., Zisimopoulos, P., Sigaras, A., Meseguer, M., Miller, K.A., Hoffman, D., Rosenwaks, Z., Elemento, O., Zaninovic, N., Hajirasouliha, I. (2023). Accurate and Noninvasive Ploidy Prediction for Human Preimplantation Embryos. [Insert journal name and volume here].