AUTHOR=Krishnan Palani Thanaraj , Krishnadoss Pradeep , Khandelwal Mukund , Gupta Devansh , Nihaal Anupoju , Kumar T. Sunil 

TITLE=Enhancing brain tumor detection in MRI with a rotation invariant Vision Transformer

JOURNAL=Frontiers in Neuroinformatics

VOLUME=Volume 18 - 2024

YEAR=2024

URL=https://www.frontiersin.org/journals/neuroinformatics/articles/10.3389/fninf.2024.1414925

DOI=10.3389/fninf.2024.1414925

ISSN=1662-5196

ABSTRACT=The Rotational Invariant Vision Transformer (RViT) is introduced in this research, a deep learning model specifically designed for the classification of brain tumors using MRI scans. RViT integrates rotated patch embeddings to improve the accuracy of brain tumor identification. When assessed on the Brain Tumor MRI Dataset from Kaggle, RViT displays superior performance in terms of sensitivity (1.0), specificity (0.975), F1-score (0.984), Matthew's Correlation Coefficient (MCC) (0.972), and attains an overall accuracy of 0.986. These findings not only surpass the standard Vision Transformer model but also outperform several techniques, signifying its effectiveness in the realm of medical imaging. The study affirms that the inclusion of rotational patch embeddings enhances the model's ability to handle diverse orientations, a common challenge in tumor imaging. Our results show that the specialized architecture and rotational invariance approach of RViT could enhance existing methodologies for brain tumor detection, with potential expansions to other intricate imaging tasks.