A Deep Learning Approach for Masking Fetal Gender in Ultrasound Images
This addresses the ethical issue of sex-selective abortion in prenatal care by improving ultrasound technology accessibility, though it is incremental as it applies an existing method (YOLOv5L) to a new domain.
The paper tackles the problem of fetal gender determination in ultrasound images leading to sex-selective abortion by proposing a deep learning object detection approach to mask fetal gender, achieving 45.8% AP[0.5:0.95], 92% F1-score, and an 85% reduction in false negative rate with a bounding box delay rule.
Ultrasound (US) imaging is highly effective with regards to both cost and versatility in real-time diagnosis; however, determination of fetal gender by US scan in the early stages of pregnancy is also a cause of sex-selective abortion. This work proposes a deep learning object detection approach to accurately mask fetal gender in US images in order to increase the accessibility of the technology. We demonstrate how the YOLOv5L architecture exhibits superior performance relative to other object detection models on this task. Our model achieves 45.8% AP[0.5:0.95], 92% F1-score and 0.006 False Positive Per Image rate on our test set. Furthermore, we introduce a bounding box delay rule based on frame-to-frame structural similarity to reduce the false negative rate by 85%, further improving masking reliability.