当前位置: X-MOL 学术Biomed. Eng. Biomed. Tech. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
A robust grey wolf-based deep learning for brain tumour detection in MR images.
Biomedical Engineering / Biomedizinische Technik ( IF 1.7 ) Pub Date : 2020-04-28 , DOI: 10.1515/bmt-2018-0244
A Geetha 1 , N Gomathi 2
Affiliation  

In recent times, the detection of brain tumours has become more common. Generally, a brain tumour is an abnormal mass of tissue where the cells grow uncontrollably and are apparently unregulated by the mechanisms that control cells. A number of techniques have been developed thus far; however, the time needed in a detecting brain tumour is still a challenge in the field of image processing. This article proposes a new accurate detection model. The model includes certain processes such as preprocessing, segmentation, feature extraction and classification. Particularly, two extreme processes such as contrast enhancement and skull stripping are processed under the initial phase. In the segmentation process, we used the fuzzy means clustering (FCM) algorithm. Both the grey co-occurrence matrix (GLCM) as well as the grey-level run-length matrix (GRLM) features were extracted in the feature extraction phase. Moreover, this paper uses a deep belief network (DBN) for classification. The optimized DBN concept is used here, for which grey wolf optimisation (GWO) is used. The proposed model is termed the GW-DBN model. The proposed model compares its performance over other conventional methods in terms of accuracy, specificity, sensitivity, precision, negative predictive value (NPV), the F1Score and Matthews correlation coefficient (MCC), false negative rate (FNR), false positive rate (FPR) and false discovery rate (FDR), and proves the superiority of the proposed work.

中文翻译:

基于强灰狼的深度学习,可用于MR图像中的脑肿瘤检测。

近年来,脑肿瘤的检测变得越来越普遍。通常,脑瘤是组织的异常肿块,其中细胞不受控制地生长,并且显然不受控制细胞的机制的调节。到目前为止,已经开发了许多技术。然而,在图像处理领域,检测脑肿瘤所需的时间仍然是一个挑战。本文提出了一种新的精确检测模型。该模型包括某些过程,例如预处理,分割,特征提取和分类。特别是,在初始阶段要处理两个极端过程,例如对比度增强和头骨剥离。在分割过程中,我们使用了模糊均值聚类(FCM)算法。在特征提取阶段同时提取了灰色共现矩阵(GLCM)和灰度级运行长度矩阵(GRLM)特征。此外,本文使用深度信念网络(DBN)进行分类。这里使用了优化的DBN概念,为此使用了灰太狼优化(GWO)。提出的模型称为GW-DBN模型。提出的模型在准确性,特异性,敏感性,精密度,阴性预测值(NPV),F1Score和Matthews相关系数(MCC),假阴性率(FNR),假阳性率(FPR)方面与其他常规方法进行了比较)和错误发现率(FDR),并证明了拟议工作的优越性。这里使用了优化的DBN概念,为此使用了灰太狼优化(GWO)。提出的模型称为GW-DBN模型。提出的模型在准确性,特异性,敏感性,精密度,阴性预测值(NPV),F1Score和Matthews相关系数(MCC),假阴性率(FNR),假阳性率(FPR)方面与其他常规方法进行了比较)和错误发现率(FDR),并证明了拟议工作的优越性。这里使用了优化的DBN概念,为此使用了灰太狼优化(GWO)。提出的模型称为GW-DBN模型。提出的模型在准确性,特异性,敏感性,精密度,阴性预测值(NPV),F1Score和Matthews相关系数(MCC),假阴性率(FNR),假阳性率(FPR)方面与其他常规方法进行了比较)和错误发现率(FDR),并证明了拟议工作的优越性。
更新日期:2020-04-28
down
wechat
bug