Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2501.15248 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915369877241856 |
|---|---|
| author | Tian, Yueying Ucurum, Elif Han, Xudong Young, Rupert Chatwin, Chris Birch, Philip |
| author_facet | Tian, Yueying Ucurum, Elif Han, Xudong Young, Rupert Chatwin, Chris Birch, Philip |
| contents | Ultrasound imaging is widely used in medical diagnosis, especially for fetal health assessment. However, the availability of high-quality annotated ultrasound images is limited, which restricts the training of machine learning models. In this paper, we investigate the use of diffusion models to generate synthetic ultrasound images to improve the performance on fetal plane classification. We train different classifiers first on synthetic images and then fine-tune them with real images. Extensive experimental results demonstrate that incorporating generated images into training pipelines leads to better classification accuracy than training with real images alone. The findings suggest that generating synthetic data using diffusion models can be a valuable tool in overcoming the challenges of data scarcity in ultrasound medical imaging. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2501_15248 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Enhancing Fetal Plane Classification Accuracy with Data Augmentation Using Diffusion Models Tian, Yueying Ucurum, Elif Han, Xudong Young, Rupert Chatwin, Chris Birch, Philip Computer Vision and Pattern Recognition Ultrasound imaging is widely used in medical diagnosis, especially for fetal health assessment. However, the availability of high-quality annotated ultrasound images is limited, which restricts the training of machine learning models. In this paper, we investigate the use of diffusion models to generate synthetic ultrasound images to improve the performance on fetal plane classification. We train different classifiers first on synthetic images and then fine-tune them with real images. Extensive experimental results demonstrate that incorporating generated images into training pipelines leads to better classification accuracy than training with real images alone. The findings suggest that generating synthetic data using diffusion models can be a valuable tool in overcoming the challenges of data scarcity in ultrasound medical imaging. |
| title | Enhancing Fetal Plane Classification Accuracy with Data Augmentation Using Diffusion Models |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2501.15248 |