International Journal of Medical Engineering and Informatics · 2015 · 16 citations · 19 references
ComputingComputer ScienceRadiologyDiagnostic ImagingDiagnostic RadiologyMachine LearningComputed TomographyComputer-aided DiagnosisImage ClassificationMedical Image AnalysisMedical ImagingPattern RecognitionClassification AccuracyClassification Accuracy AnalysisClassification Accuracy AssessmentFuzzy Inference SystemsFuzzy Logic ClassificationGray-level Co-occurrence MatricesLung Tumor ClassificationMedical Image Classification
In this paper a computer aided classification method for computed tomography (CT) images of lungs using fuzzy inference system (FIS) and adaptive neuro fuzzy inference system (ANFIS) is proposed. The entire lung lobe is segmented from the CT images using morphological operations. Statistical and gray level co-occurrence matrix (GLCM) parameters are calculated from the segmented image. Among 14 GLCM parameters and three statistical parameters, four parameters are selected for classification by principal component analysis. The parameters selected are cluster shade, dissimilarity, difference variance and skewness. The classification process is done by FIS and ANFIS. Compared to FIS, ANFIS gives better classification. A new training algorithm is proposed for the back propagation neural network used in the ANFIS. The proposed method gives a classification accuracy of 94% with a specificity of 100% and accuracy of 93%.
19
Samuel G. Armato, Geoffrey McLennan, Luc Bidaut et al. · Medical Physics · 2011 · 2.8K citations · Full text
Screening for Early Lung Cancer with Low-Dose Spiral CT: Prevalence in 817 Asymptomatic Smokers
Stefan Diederich, Dag Wormanns, M Semik et al. · Radiology · 2002 · 506 citations
Feature Selection Using Principal Component Analysis
Fengxi Song, Zhongwei Guo, Dayong Mei · 2010 · 386 citations