AUTHOR=Sarah Ponuku , Krishnapriya Srigiri , Saladi Saritha , Karuna Yepuganti , Bavirisetti Durga Prasad 

TITLE=A novel approach to brain tumor detection using K-Means++, SGLDM, ResNet50, and synthetic data augmentation

JOURNAL=Frontiers in Physiology

VOLUME=Volume 15 - 2024

YEAR=2024

URL=https://www.frontiersin.org/journals/physiology/articles/10.3389/fphys.2024.1342572

DOI=10.3389/fphys.2024.1342572

ISSN=1664-042X

ABSTRACT=Brain tumors are abnormal growth of cells that occur when cells around the brain multiply out of control. The therapeutic management of brain tumors continues to be a major problem due to the distinct genetic, epigenetic, and microenvironmental characteristics of neural tissues that impart treatment resistance. The challenge this research attempts to solve is the precise and reliable early detection of tumors in magnetic resonance imaging (MRI) images. The main goal of this work is to discover the finest deep-learning model for classifying tumors from brain MRI data. The proposed method combines segmentation using K-means++, feature extraction from Spatial gray level dependence matrix (SGLDM), classification using ResNet50, and synthetic data augmentation. In terms of accuracy, sensitivity, and specificity, testing on the Br35H::Brain Tumor Detection 2020 dataset revealed that the suggested method outperforms existing state-of-the-art methods. The findings of this work show the efficiency of deep learning models in healthcare and how they might help doctors make quick, precise decisions.