2024-12-23-varma24a.md

title	abstract	layout	series	publisher	issn	id	month	tex_title	firstpage	lastpage	page	order	cycles	bibtex_author	author	date	address	container-title	volume	genre	issued	pdf	extras
VariViT: A Vision Transformer for Variable Image Sizes	Vision Transformers (ViTs) have emerged as the state-of-the-art architecture in representation learning, leveraging self-attention mechanisms to excel in various tasks. ViTs split images into fixed-size patches, constraining them to a predefined size and necessitating pre-processing steps like resizing, padding, or cropping. This poses challenges in medical imaging, particularly with irregularly shaped structures like tumors. A fixed bounding box crop size produces input images with highly variable foreground-to-background ratios. Resizing medical images can degrade information and introduce artifacts, impacting diagnosis. Hence, tailoring variable-sized crops to regions of interest can enhance feature representation capabilities. Moreover, large images are computationally expensive, and smaller sizes risk information loss, presenting a computation-accuracy tradeoff. We propose VariViT, an improved ViT model crafted to handle variable image sizes while maintaining a consistent patch size. VariViT employs a novel positional embedding resizing scheme for a variable number of patches. We also implement a new batching strategy within VariViT to reduce computational complexity, resulting in faster training and inference times. In our evaluations on two 3D brain MRI datasets, VariViT surpasses vanilla ViTs and ResNet in glioma genotype prediction and brain tumor classification. It achieves F1-scores of 75.5% and 76.3%, respectively, learning more discriminative features. Our proposed batching strategy reduces computation time by up to 30% compared to conventional architectures. These findings underscore the efficacy of VariViT in image representation learning.	inproceedings	Proceedings of Machine Learning Research	PMLR	2640-3498	varma24a	0	VariViT: A Vision Transformer for Variable Image Sizes	1571	1583	1571-1583	1571	false	Varma, Aswathi and Shit, Suprosanna and Prabhakar, Chinmay and Scholz, Daniel and Li, Hongwei Bran and Menze, Bjoern and Rueckert, Daniel and Wiestler, Benedikt	given family Aswathi Varma given family Suprosanna Shit given family Chinmay Prabhakar given family Daniel Scholz given family Hongwei Bran Li given family Bjoern Menze given family Daniel Rueckert given family Benedikt Wiestler	2024-12-23		Proceedings of The 7nd International Conference on Medical Imaging with Deep Learning	250	inproceedings	date-parts 2024 12 23	https://raw.githubusercontent.com/mlresearch/v250/main/assets/varma24a/varma24a.pdf