A brain tumor is considered one of the deadliest forms among all types of cancer due to its aggressive nature leading to patients’ low survival rate. Detection and classification of brain tumors have a significant impact on treatment planning and patient survival. The significance and importance of this work lie in the formulation of several probabilistic features that represent higher-level probabilistic uncertainty. To create these features, the gain function of probabilistic Hanman transform is replaced with the gain functions of Shannon, Renyi, and Tsallis entropy functions thus paving a way to the corresponding hybrid transforms, Hanman-Shannon, Hanman-Renyi, and Hanman-Tsallis transforms. The new features are extracted from brain MR images to detect and classify tumor by developing the possibilistic Hanman-Shannon transform classifier. This uses the t-normed errors between the training and testing features. The proposed system when evaluated on the two Brain MRI datasets yields the highest accuracy of around 99%.

中文翻译：

---

使用概率特征和可能的汉曼-香农变换分类器从磁共振图像中分类脑肿瘤

脑肿瘤被认为是所有类型癌症中最致命的形式之一，因为它具有侵袭性，导致患者的存活率低。脑肿瘤的检测和分类对治疗计划和患者生存有重大影响。这项工作的意义和重要性在于提出了几个代表更高级别概率不确定性的概率特征。为了创建这些特征，概率 Hanman 变换的增益函数被 Shannon、Renyi 和 Tsallis 熵函数的增益函数取代，从而为相应的混合变换 Hanman-Shannon、Hanman-Renyi 和 Hanman-Tsallis 变换铺平了道路. 通过开发可能的 Hanman-Shannon 变换分类器，从脑 MR 图像中提取新特征以检测和分类肿瘤。这使用了训练和测试特征之间的 t 范数误差。在对两个脑 MRI 数据集进行评估时，所提出的系统产生了大约 99% 的最高准确度。