Automated computer-aided soft computing methods are presently used to detect the tumor regions in brain images. In this paper, the tumor cells are detected in the brain Magnetic Resonance Imaging (MRI) using the Extended Linear Boosting (ELB) classification method as one type of soft computing process. This paper proposes an effective brain tumor detection and segmentation method using the ELB classification method. The Curvelet transform is applied on the source brain MRI image to convert the spatial domain pixels into multi-resolution pixel. The spectral and linear discriminate features are computed from the Curvelet transformed coefficient matrix. The dimensionality of the computed features is reduced using the PCA method and the optimized features are then classified using the ELB classification method. The performance evaluation metrics, sensitivity, specificity, accuracy and detection rate, are used in this paper to evaluate the performance of the proposed brain tumor detection and segmentation system.

自动化计算机辅助软计算方法目前用于检测脑图像中的肿瘤区域。在本文中，使用扩展线性增强 (ELB) 分类方法作为一种软计算过程，在脑磁共振成像 (MRI) 中检测肿瘤细胞。本文提出了一种使用ELB分类方法的有效脑肿瘤检测和分割方法。Curvelet 变换应用于源脑 MRI 图像，将空间域像素转换为多分辨率像素。光谱和线性判别特征是根据 Curvelet 变换系数矩阵计算的。使用 PCA 方法降低计算特征的维数，然后使用 ELB 分类方法对优化后的特征进行分类。绩效评估指标，