Abstract:
Brain tumors are a leading cause of mortality in India, with over 28,000 cases reported annually,
resulting in more than 24,000 deaths per year, according to the International Association of Cancer Registries.
Early detection, segmentation, and accurate classification are crucial in effective tumor analysis, and various
algorithms have been developed to achieve this. In this study, we propose a novel approach for the detection and
classification of brain tumors using both single slices of MRI and CT, as well as input-level fused images of
MRI & CT. Our approach involves the implementation of the PrinciResnet brain tumor classification technique,
which is based on Principal Component Analysis (PCA) and Resnet techniques. We report that our approach
significantly improves the accuracy, sensitivity, and specificity parameters to 90%, 96%, and 95%, respectively,
based on a dataset of 600 fused slices and 1000 single slices obtained from reputable sources. Our findings hold
promise for improving the diagnosis and treatment of brain tumors, which are a significant cause of mortality
globally.
