Retracted: Comparison of Accuracy and Sensitivity in Liver Cancer Segmentation of Magnetic Resonance Images using Convolutional Neural Network in Comparison with Support Vector Machine
S. Charan, G. Uganya, M. Naveen Kumar
Abstract
The main aim of this work is to detect liver cancer by segmentation of magnetic resonance images using Convolutional Neural Network in comparison of accuracy and sensitivity with Support Vector Machine Classifier. Materials and Methods: Group 1 was taken as Convolutional Neural Network and Group 2 was taken as Support Vector Machine. These groups are analyzed by an independent sample T-test with a pretest power of 80% whereas total number of samples N=40. Results: Convolutional Neural Network achieves an accuracy of 96.6025% and sensitivity of 97.61%. Support Vector Machine achieves an accuracy of 86.945% and sensitivity 94.385%. There is a significant difference of 0.036 for accuracy and 0.041 for sensitivity. Conclusion: Convolutional Neural Network achieves significantly better accuracy and sensitivity when compared with Support Vector Machine.