Enhancing Medical Image Classification with Vision Transformers on Diverse Datasets
Abstract: Medical image classification is essential for accurate diagnosis and effective treatment planning. This research investigates the implementation of MedViT, a robust Vision Transformer tailored for medical image analysis, and compares its performance against four models: StarterNet, TinyVGG, a standard Vision Transformer (ViT), and a Convolutional Neural Network (CNN). Evaluations conducted on PathMINST, a medical imaging dataset, and CIFAR - 10, a general - purpose image classification dataset, to assess model generalization.
Keywords: MedViT, Vision Transformer, Medical Image Classification, PathMINST, CIFAR-10
How to Cite?: Aditya Dhar Dwivedi, "Enhancing Medical Image Classification with Vision Transformers on Diverse Datasets", Volume 14 Issue 1, January 2025, International Journal of Science and Research (IJSR), Pages: 680-693, https://www.ijsr.net/getabstract.php?paperid=SR25113171116, DOI: https://dx.doi.org/10.21275/SR25113171116