HOD, Department of BME, Alpha College of Engineering, Chennai, 124, India.
St. Peter's College of Engineering and Technology, Avadi, Chennai, 54, India.
J Med Syst. 2019 Feb 13;43(3):77. doi: 10.1007/s10916-019-1177-9.
The Lung nodules are very important to indicate the lung cancer, and its early detection enables timely treatment and increases the survival rate of patient. Even though lots of works are done in this area, still improvement in accuracy is required for improving the survival rate of the patient. The proposed method can classify the stages of lung cancer in addition to the detection of lung nodules. There are two parts in the proposed method, the first part is used for classifying normal/abnormal and second part is used for classifying stages of lung cancer. Totally 10 features from the lung region segmented image are considered for detection and classification. The first part of the proposed method classifies the input images with the aid of Naive Bayes classifier as normal or abnormal. The second part of the system classifies the four stages of lung cancer using Neuro Fuzzy classifier with Cuckoo Search algorithm. The results of proposed system show that the rate of accuracy of classification is improved and the results are compared with SVM, Neural Network and Neuro Fuzzy Classifiers.
肺结节对于指示肺癌非常重要，早期发现可以实现及时治疗并提高患者的存活率。尽管在这一领域已经开展了大量工作，但为了提高患者的存活率，仍需要提高准确性。所提出的方法除了检测肺结节外，还可以对肺癌的分期进行分类。该方法有两部分，第一部分用于分类正常/异常，第二部分用于分类肺癌分期。总共考虑了来自肺区域分割图像的 10 个特征用于检测和分类。所提出方法的第一部分借助朴素贝叶斯分类器对输入图像进行分类，分为正常或异常。系统的第二部分使用带有布谷鸟搜索算法的神经模糊分类器对肺癌的四个阶段进行分类。所提出系统的结果表明，分类的准确性得到了提高，并与 SVM、神经网络和神经模糊分类器进行了比较。
