Enhancing Fetal Plane Classification Accuracy with Data Augmentation Using Diffusion Models

TL;DR

This paper explores using diffusion models to generate synthetic ultrasound images, enhancing fetal plane classification accuracy by augmenting limited real data, thus improving machine learning performance in medical diagnosis.

Contribution

It introduces a novel approach of applying diffusion models for synthetic data generation to improve ultrasound image classification accuracy.

Findings

Synthetic images improve classification accuracy.

Fine-tuning with real images enhances performance.

Diffusion models effectively augment ultrasound datasets.

Abstract

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.