Intracardiac Mass Detection and Classification Using Double Convolutional Neural Network Classifier
Background: Identification and classification of intracardiac masses in echocardiogram is one of the significant processes in the diagnosis of cardiovascular disease. There is some automatic detection and classification techniques for intracardiac mass are addressed in the literature to improve the classification accuracy of intracardiac mass. However, these strategies have a low false error ratio during the analysis.
Material & methods: A robust back propagation neural network (RBPNN) technique is used to conquer every single conventional-issue utilizing the echocardiogram image analysis for this work, which consists of four phases such as noise removal, automatic segmentation, feature extraction, and intracardiac masses classification. Initially, the noise is diminished from the echocardiogram images utilizing the adaptive vector median filter (AVMF). Then, linear iterative vessel segmentation (LIVS) is applied for automatic segmentation of the masses followed by the extraction of texture features using the multiscale local binary pattern (MS-LBP) approach. Finally, RBPNN is employed to classify the heart mass from the images of echocardiogram with the layered kernel for the system combination.
Results and main findings: The proposed AVMF-MS-LBP based RBPNN approach can help the radiologists to diagnose and automatically classify the heart diseases from the echocardiogram images. In addition, several medical statistical parameters like sensitivity, specificity, and accuracy are computed to evaluate the performance of proposed AVMF-MS-LBP based RBPNN approach with comparison to the existing approaches like sparse representation, Support Vector Machine (SVM), SVM with Particle Swarm Optimization (SVM-PSO), Artificial Neural Network (ANN) and Kernel Collaborative Representation (KCR).
Conclusions: Extensive simulation results obtained using proposed AVMF-MS-LBP based RBPNN approach disclosed the superiority over existing intracardiac mass detection and classification approaches in terms of accuracy, sensitivity, and specificity (i.e., 97.845%, 97.38%, and 98.31% respectively).Therefore, proposed AVMF-MS-LBP based RBPNN is more precise and prominent approach that assist the cardiologists to make the anticipation before going for medical surgery.