Using Double Convolution Neural Network for Lung Cancer Stage Detection

Abstract

Recently, deep learning is used with convolutional Neural Networks for image classification and figure recognition. In our research, we used Computed Tomography (CT) scans to train a double convolutional Deep Neural Network (CDNN) and a regular CDNN. These topologies were tested against lung cancer images to determine the Tx cancer stage in which these topologies can detect the possibility of lung cancer. The first step was to pre-classify the CT images from the initial dataset so that the training of the CDNN could be focused. Next, we built the double Convolution deep Neural Network with max pooling to perform a more thorough search. Finally, we used CT scans of different Tx cancer stages of lung cancer to determine the Tx stage in which the CDNN would detect possibility of lung cancer. We tested the regular CDNN against our double CDNN. Using this algorithm, doctors will have additional help in early lung cancer detection and early treatment. After extensive training with 100 epochs, we obtained the highest accuracy of 0.9962, whereas the regular CDNN obtained only 0.876 accuracy.

Deep Learning and Big Data in Healthcare: A Double Review for Critical Beginners

“…DL solutions have been applied to medical image analysis [120] in very different topics related to classification or detection, such as identifying lung nodules into benign or malignant [121], alcoholism identification through brain MRI [122,123], multiple sclerosis detection [124], categorizing chest X-rays into different diseases [125], distinguishing patients with Alzheimer's disease versus normal [126], diagnosing diabetic retinopathy using digital photographs of the fundus of the eye [127], discriminating kidney cancer histopathological images into tumor or non-tumor [128], detecting cancerous lung nodules [129] and lung cancer stage [130] on CT scans, malignant skin cells on dermatological photographs [131], mitotic figures in breast histology images [132], or cell nuclei in colorectal adenocarcinoma histology image [133]. Regarding image segmentation, DL covers a variety of organs such as liver, prostate, spine, and knee cartilage both in CT and MRI [116,134].…”

The Performance of Deep Learning Algorithms on Automatic Pulmonary Nodule Detection and Classification Tested on Different Datasets That Are Not Derived from LIDC-IDRI: A Systematic Review

“…Unless otherwise stated, the CT images used in the training dataset were not a part of the test dataset. [22] 2019 87.5 Ciompi, Francesco et al [29] 2017 79.5 * Jakimovski, Goran et al [30] 2019 99.6 Lakshmanaprabu, S.K. et al [31] 2018 94.5 Liao, Fangzhou et al [23] 2019 81.4 Liu, Xinglong et al [33] 2017 90.3 * Masood, Anum et al [21] 2018 96.3 Nishio, Mizuho et al [34] 2018 68 Onishi, Yuya et al [35] 2018 81.7 Polat, Huseyin et al [36] 2019 91.8 Qiang, Yan et al [37] 2017 82.8 Rangaswamy et al [38] 2019 96 Sori, Worku Jifara et al [39] 2018 87.8 Wang, Shengping et al [40] 2018 84 Wang, Yang et al [25] 2019 87.3 Yuan, Jingjing et al [41] 2017 93.9 * Zhang, Chao et al [42] 2019 92 * (c)…”

“…Jakimovski and Davcev [30] used an algorithm that was both trained and tested on the Image and Data Archive of the University of South Carolina and Laboratory of Neuro Imaging (LONI database) [48] and achieved an accuracy of 99.6%, a sensitivity of 99.9%, and specificity of 98.6% for their best-performing algorithm. The algorithm from Jakimovski et al [30] outputted a single decimal value between 0.0 and 1.0, where 0.0 was not cancer and 1.0 was cancer. They converted the value to a percentage and set a minimal threshold value at 73% before the image was categorized as cancer.…”

Lung Tumor Segmentation and Staging from CT Images Using Fast and Robust Fuzzy C-Means Clustering

“…Many different works have been done and reported in literature previously to detect and classify the lung tumor from CT image using various types of algorithms [3][4][5][6][7][8][9][10][11][12][13][14]. Lung tumor is often segmented manually which is user and experience dependent, subjective, and may lead to erroneous diagnosis.…”

“…The performance of these algorithms declines without enough training data. Recently deep neural networks are becoming popular to address classification problems [9,10]. But large training datasets and huge computational cost are needed for deep learning.…”