Diagnosis of Brain Tumor using Semantic Segmentation and Advance-CNN Classification

The brain cancer prediction is mediocre at an early stage as it is impotent by the radiologist. Various investigations done so far manifest clearly that the nodule segmentation algorithms are ineffectual. Thus, this investigation has centralized semantic segmentation based Nano segmentation method for precise segmentation of lesion. The supreme intent of this research paper is the enhancement of brain MRI images to recognize the tumor efficiently and small-scale anomalous nodules segmentation in the brain region. The initial step Isolateral filter enhancement techniques which can eradicate the noise discern in the images. In the subsequent step the semantic Segmentation Based Nano area detection algorithm is implemented in an enhanced nodule image sequence for abnormal brain tissue prediction. Ultimately, the brain nodule images are procured by utilizing a deep learning based Advanced CNN (ACNN). Nano Segmentation method and the deep learning classification (DLC) method has an accuracy of 95.7% that helps to diagnose the cancer cells using the feature extraction process which is done automatically. Average segmentation time for nodule slice order is 1.01s. Comparative analysis is made with ResNet-50 based on the different testing and training data at the rate of 90% -10%, 80%-20% and 70%-30% respectively which proves the robustness of the proposed research work. Experimental results prove the proposed system effectiveness when compared with other detection methods.