A Novel Architecture for Predicting Pneumonia Patients by using LSTM, GRU and CNN
Nitin Arora1, Ahatsham2, Anupam Singh3, Vivek Shahare4
1Nitin Arora*, School of Computer Science, University of Petroleum and Energy Studies, Dehradun, India.
2Ahatsham, School of Computer Science, University of Petroleum and Energy Studies, Dehradun, India.
3Anupam Singh, School of Computer Science, University of Petroleum and Energy Studies, Dehradun, India.
4Vivek Shahare, School of Computer Science, University of Petroleum and Energy Studies, Dehradun, India.
Manuscript received on September 10, 2019. | Revised Manuscript received on October 20, 2019. | Manuscript published on October 30, 2019. | PP: 4120-4126 | Volume-9 Issue-1, October 2019 | Retrieval Number: A1353109119/2019©BEIESP | DOI: 10.35940/ijeat.A1353.109119
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: The Models based only on Long-Short Term Memory (LSTM), Gated Recurrent Unit (GRU) and Dynamic Recurrent Neural Network (Dynamic RNN) are not sufficient for prediction of a pneumonia patients using image processing. The proposed network uses the properties of LSTM, GRU and Convolutional Neural Network like capacity to remember long-term memory and handling the input parameters dynamically. Investigating these networks, it is found that the proposed network have a deep insight on the medical image before it can remember then forward the necessary information. In the proposed model, the LSTM and GRU directly connected with dual 1×1 convolutional network followed by the classification layers resulting with better performance. The proposed model provides the test accuracy of 94.20% and test loss of 0.04749 when tested on dataset of pneumonia patients consisting of 2 classes (normal chest, diseased chest with pneumonia) has provided better results than LSTM, GRU, Dynamic RNN and Convolutional with LSTM in all aspects like training loss, training accuracy, testing loss and testing accuracy.
Keywords: LSTM, GRU, DRNN, LRCN, 1DCRNN.