Advanced CNN-Based Brain Tumor Detection and Segmentation Using MATLAB: A Diagnostic Accuracy Study

Abstract

The brain tumor is intracranial mass made up by abnormal growth of tissue in the brain or around the brain. The new CNN technique in MATLAB can be used to detect the brain tumors more precisely. The study objective was to get the easily diagnose of the Tumor and to get the Segmentation of the Tumor using MATLAB. This analytical study was conducted in MATLAB which is a software. The sample size of 100 was estimated by using 95% confidence level with 5% margin of error and taking an expected percentage of Brain Tumor as 99%. The data consisted of 100 patients. The data was gathered from both males and females. There was total 97 patients, out of which around 67 were males and around 33 were female. According to the previous method diagnosis, there were around 42 patients (42%) that had the benign indication of Brain Tumor and around 58 patients (58%) that malignant indication of Brain Tumor. According to MATLAB results, around 51 patients (51%) had benign type of Brain Tumor and around 49 patients (49%) had malignant type of cervical lymphadenopathy. The main analytical statistics proved the diagnostic odd ratio statistics to be around 97% with the sensitivity of MATLAB in differentiating between benign and malignant Brain Tumor was around 99% with the specificity around 98% with disease prevalence to be around 96% with positive and negative predictive values to be around 99% and 1%. Our experimental results demonstrated that both models enhance the prediction performance of diagnosis of brain tumors. We achieved 97.8% and 100% prediction accuracy for dataset 1 and dataset 2, respectively outperforming previous studies found in the literature. Therefore, we believe that our proposed methods are outstanding candidates for brain tumor detection.