Abstract:Purpose: To explore the feasibility of classifying hepatocellular carcinoma (HCC) and hepatic hemangioma (HEM) using texture features of non-enhanced computed tomography (CT) images, especially to investigate the effectiveness of a novel texture analysis method based on the combination of wavelet and co-occurrence matrix. Methods: 269 patients were retrospectively analyzed, including 129 HCCs and 140 HEMs. We cropped tumor regions of interest (ROIs) on non-enhanced CT images, and then used four texture analysis methods to extract quantitative data of the ROIs: gray-level histogram (GLH), gray-level co-occurrence matrix (GLCM), reverse biorthogonal wavelet transform (RBWT), and reverse biorthogonal wavelet co-occurrence matrix (RBCM). The RBCM was a novel method proposed in this study that combined wavelet transform and co-occurrence matrix. It discretized wavelet coefficient matrices based on the statistical characteristics of the training set. Thus, four sets of texture features were obtained. We then conducted classification studies using support vector machine on each set of texture features. 10-fold cross training and testing were used in the classifications, and their results were evaluated and compared. In addition, we tested the significant differences in the texture features of the RBCM method and explored the possible relationships between textures and lesion types. Results: The RBCM method achieved the best classification performance: its average accuracy was 82.14%; its average AUC (area under the receiver operating characteristic curve) was 0.8423. In addition, using the methods of GLH, GLCM, and RBWT, their average accuracies were 75.81%, 78.79%, and 78.8%, respectively. Conclusions: It indicates that the developed texture analysis methods are rewarding for computer-aided diagnosis of HCC and HEM based on non-enhanced CT images. Furthermore, the distinguishing ability of the proposed RBCM method is more pronounced.

Classification of Hepatic Tissues from CT Images Based on Texture Features and Multiclass Support Vector Machines

Texture-specific Bag of Visual Words Model and Spatial Cone Matching-Based Method for the Retrieval of Focal Liver Lesions Using Multiphase Contrast-Enhanced CT Images

Computerized Features for LI-RADS Based Computer-Aided Diagnosis of Liver Lesions

Residual Convolutional Neural Networks With Global And Local Pathways For Classification Of Focal Liver Lesions

A Retrieval System For 3d Multi-Phase Contrast-Enhanced Ct Images Of Focal Liver Lesions Based On Combined Bags Of Visual Words And Texture Words

Multi-Phase Ct Image Based Hepatic Lesion Diagnosis By Svm

A Novel Texture Analysis Method Based on Reverse Biorthogonal Wavelet and Co-Occurrence Matrix Applied for Classification of Hepatocellular Carcinoma and Hepatic Hemangioma

An Automatic Grading Method of Liver Cirrhosis from Abdominal CT Images

Liver cancer classification via deep hybrid model from CT image with improved texture feature set and fuzzy clustering based segmentation

Texture analysis and classification of ultrasound liver images.

Classification and Prognosis Prediction from Histopathological Images of Hepatocellular Carcinoma by a Fully Automated Pipeline Based on Machine Learning

Key Techniques Research in Computer-Aided Hepatic Lesion Diagnosis System Based on Multi-Phase CT Images

Computer Vision Approach for Liver Tumor Classification Using CT Dataset

A Hybrid Approach to Diagnosis of Hepatic Tumors in Computed Tomography Images

Study of Texture Segmentation and Classification for Grading Small Hepatocellular Carcinoma Based on CT Images

Textural Features Extraction for Liver Cancer Identification Based on CT Images

The method and efficacy of support vector machine classifiers based on texture features and multi-resolution histogram from (18)F-FDG PET-CT images for the evaluation of mediastinal lymph nodes in patients with lung cancer.

Intelligent diagnosis of liver cancer based on CT

An identification method of malignant and benign liver tumors from ultrasonography based on GLCM texture features and fuzzy SVM

Automatic segmentation and classification of liver tumor from CT image using feature difference and SVM based classifier-soft computing technique

Classification of Hepatocellular Carcinoma and Intrahepatic Cholangiocarcinoma Based on Radiomic Analysis.