Differential Region-Based Segmentation Method for Brain Tumor Detection using Computed Tomography Images

Brain malignant growth detection is eased using medical images and machine learning paradigms. This article introduces a differential region-based segmentation method to improve brain tumor detection precision from computerized tomography inputs. The proposed method relies on neural networks for different region detection and segmentation. The neural network is trained using the different connecting regional conditions. Based on the learning iterations the region matching conditions are verified to ensure maximum segments are identified. Thus this proposed method improves accuracy and precision by 15.48% and 12.59% for the different features.