Answers are grounded in this paper and cite their sources.
DICaugment is a Python package based on the popular image augmentation library Albumentations (Buslaev et al., 2020), with specific enhancements for working with volumetric medical images, such as Computed Tomography (CT) scans.This package provides a collection of over 40 powerful augmentation techniques that can be seamlessly integrated into a machine-learning pipeline to augment volumetric medical images.The package was designed to incorporate the image acquisition metadata available in DICOM headers, allowing users to create transformations that are consistent with those acquisition parameters for specific CT systems and reconstruction kernels.DICaugment extends the success of the Albumentations library for two-dimensional (2D) image augmentation to the realm of three-dimensional (3D) images, offering a comprehensive set of transformations and augmentations, ranging from pixel-level intensity transformations to spatial transformations, all designed and optimized for 3D data.
No takes yet. Share an insight, caveat, or question.
McIntosh et al. (2024) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: