DEEP NEURAL NETWORK BASED CLASSIFICATION MODEL FOR FEATURE EXTRACTION
European Journal of Molecular & Clinical Medicine,
2020, Volume 7, Issue 10, Pages 3539-3554
Abstract
Background:Computer aided diagnosis model developed for the detection of lung cancer can leads to higher survival rate. In recent days, Deep Learning (DL) models finds successful for various applications in medical domain.
Materials and Methods:
In this paper, an effective feature extraction based lung cancer classification model has been presented. The proposed model makes use of Hybrid Feature Extraction (HFE) with Deep Neural Network (DNN) based classification models. The proposed HFE-DNN model undergoes pre-processing to enhance the image quality. Then, feature extraction process is carried out by the use of HFE technique followed by DNN based data classification. The outcome of the HFE-DNN technique can assist doctors in the diagnosis process of lung cancer.
Results and Conclusion:
An extensive simulation results takes place on the benchmark HFE-DNN model and the results confirmed its superior characteristics over the existing methods. The simulation outcome pointed out that the proposed HFE-DNN model has offered effective classification outcome with the maximum accuracy of 85.79%, sensitivity of 79.55% and specificity of 89.03%.
Keywords:
- Article View: 124
- PDF Download: 337