Gunasundari C, Selva Bhuvaneswari K
School of Computing, SRM Institute of Science and Technology, Tiruchirappalli, India.
Department of Computer Science and Engineering, University College of Engineering Kancheepuram, Kancheepuram, India.
Sci Rep. 2025 Mar 10;15(1):8252. doi: 10.1038/s41598-025-87934-4.
A brain tumor is regarded as one of the deadliest types of cancer due to its intricate nature.This is why it is important that patients get the best possible diagnosis and treatment options. With the help of machine vision, neurologists can now perform a more accurate and faster diagnosis. There are currently no suitable methods that can be used to perform brain segmentation using image processing recently neural network model is used that it can perform better than other methods. Unfortunately, due to the complexity of the model, performing accurate brain segmentation in real images is not feasible. The main objective is to develop a novel method that can be used to analyze brain tumors using a component analysis. The proposed model consists of a deep neural network and an image processing framework. It is divided into various phases, such as the mapping stage, the data augmentation stage, and the tumor discovery stage. The data augmentation stage involves training a CNN to identify the regions of the image that are overlapping with the tumor space marker. The DCNN's predicted performance is compared with the test result. The third stage is focused on training a deep neural system and a SVM. This model was able to achieve a 99% accuracy rate and a sensitivity of 0.973%. It is primarily utilized for identifying brain tumors.
脑肿瘤因其复杂的性质被视为最致命的癌症类型之一。这就是为什么患者获得尽可能最佳的诊断和治疗方案很重要的原因。借助机器视觉，神经科医生现在可以进行更准确、更快速的诊断。目前没有合适的方法可用于使用图像处理进行脑部分割，最近使用了神经网络模型，它比其他方法表现得更好。不幸的是，由于模型的复杂性，在真实图像中进行精确的脑部分割是不可行的。主要目标是开发一种可用于通过成分分析来分析脑肿瘤的新方法。所提出的模型由一个深度神经网络和一个图像处理框架组成。它分为各个阶段，如映射阶段、数据增强阶段和肿瘤发现阶段。数据增强阶段涉及训练一个卷积神经网络（CNN）来识别图像中与肿瘤空间标记重叠的区域。将深度卷积神经网络（DCNN）的预测性能与测试结果进行比较。第三阶段专注于训练一个深度神经系统和一个支持向量机（SVM）。该模型能够达到99%的准确率和0.973%的灵敏度。它主要用于识别脑肿瘤。
