sanoscience / babynet

MICCAI2022: BabyNet: Residual Transformer Module for Birth Weight Prediction on Fetal Ultrasound Video

BabyNet: Residual Transformer Module for Birth Weight Prediction on Fetal Ultrasound Video

This is the official released code for "BabyNet: Residual Transformer Module for Birth Weight Prediction on Fetal Ultrasound Video" early accepted for the 25th International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI) 2022 in Singapore.

Abstract

Predicting fetal weight at birth is an important aspect of perinatal care, particularly in the context of antenatal management, which includes the planned timing and mode of delivery. Accurate prediction of weight using prenatal ultrasound is challenging as it requires images of specific fetal body parts during advanced pregnancy - this, however, is complicated by the poor quality of images caused by the lack of amniotic fluid. It follows that predictions which rely standard methods often suffer from significant errors. In this paper we propose the Residual Transformer Module, that extends a 3D ResNet-based network for analysis of 2D+t spatio-temporal ultrasound video scans. Our end-to-end method, called BabyNet, fully automatically predicts fetal birth weight based on fetal ultrasound video scans. We evaluate BabyNet using a dedicated clinical set comprising 225 2D fetal ultrasound videos of pregnancies from 75 patients performed one day prior to delivery. Experimental results show that BabyNet outperforms several state-of-the-art methods and estimate the weight at birth with accuracy comparable to human experts. Furthermore, combining estimates provided by human experts with those computed by BabyNet yields the best results, outperforming either method by a significant margin.

Usage

Train the model

python3 train_video.py

If you are using our codes, please cite our work:

@inproceedings{plotka2022babynet,
  title={BabyNet: Residual Transformer Module for Birth Weight Prediction on Fetal Ultrasound Video},
  author={P{\l}otka, Szymon and Grzeszczyk, Michal K and Brawura-Biskupski-Samaha, Robert and Gutaj, Pawe{\l} and Lipa, Micha{\l} and Trzci{\'n}ski, Tomasz and Sitek, Arkadiusz},
  booktitle={International Conference on Medical Image Computing and Computer-Assisted Intervention},
  pages={350--359},
  year={2022},
  organization={Springer}
}

babynet's Issues

About the dataset used in your paper.

First of all, your work is amazing! I'm very interested in the fetal birth weight prediction task you proposed. Where can I find the video dataset mentioned in your paper? Is it possiable to apply for an access to your dataset?