Revisiting Data Augmentation for Ultrasound Images
This work addresses a gap in understanding augmentation efficacy for medical imaging, specifically ultrasound, by providing a standardized benchmark and structured methodology, though it is incremental as it applies existing methods to a new domain.
The paper tackled the underutilization of data augmentation in ultrasound image analysis by analyzing the effectiveness of various techniques across 14 tasks, finding that many augmentations from natural images work well, sometimes better than ultrasound-specific ones, with TrivialAugment showing particular efficacy.
Data augmentation is a widely used and effective technique to improve the generalization performance of deep neural networks. Yet, despite often facing limited data availability when working with medical images, it is frequently underutilized. This appears to come from a gap in our collective understanding of the efficacy of different augmentation techniques across different tasks and modalities. One modality where this is especially true is ultrasound imaging. This work addresses this gap by analyzing the effectiveness of different augmentation techniques at improving model performance across a wide range of ultrasound image analysis tasks. To achieve this, we introduce a new standardized benchmark of 14 ultrasound image classification and semantic segmentation tasks from 10 different sources and covering 11 body regions. Our results demonstrate that many of the augmentations commonly used for tasks on natural images are also effective on ultrasound images, even more so than augmentations developed specifically for ultrasound images in some cases. We also show that diverse augmentation using TrivialAugment, which is widely used for natural images, is also effective for ultrasound images. Moreover, our proposed methodology represents a structured approach for assessing various data augmentations that can be applied to other contexts and modalities.