Volume - 13 | Issue-1
Volume - 13 | Issue-1
Volume - 13 | Issue-1
Volume - 13 | Issue-1
Volume - 13 | Issue-1
The proposed work extracts the tumor portion using an image fusion technique from a Multimodal MRI of Human brain tumor images. The proposed work fused the MRI modalities Flair, T2, and T1C to enhance the tumor portion from other surrounding portions in the brain slice. Preprocessing of brain tumor images are done by pseudo coloring. Modalities such as Flair, T2, and T1C are assigned three color channels, red, green, and blue, respectively and then these colored images are fused together. This enhances the tumor portion with better contrast. Gray scaletransformation is applied with a fused color image for further enhancement of the tumor portion and thus separates them from the surroundings. Finally, the tumor portion is extracted by applying Largest Connected Component (LCC) algorithm. Experiments are done using BRATS 2019 dataset and estimated with entropy (EN), structural similarity index metric (SSIM), peak-signal-to-noise ratio (PSNR), mean squared error (MSE), and dice coefficient (DC), which evaluate thesimilarity between ground truth and extracted whole tumor. Better results are produced for the proposed image than the existing method.