Pulmonary nodule diagnosis using dual‐modal supervised autoencoder based on extreme learning machine

Abstract

In recent years, deep learning techniques have been applied to the diagnosis of pulmonary nodules. In order to improve the pulmonary nodule diagnostic performance effectively, we propose a novel pulmonary nodule diagnosis method using dual‐modal deep supervised autoencoder based on extreme learning machine for which discriminative features are automatically learnt from the input data. The network is fed with nodule images in pairs obtained from computed tomography and positron emission tomography respectively. For each pair image, the high‐level discriminative features of nodules in computed tomography and positron emission tomography are extracted from stacked supervised autoencoder layers. The outputs of the proposed architecture are combined using an ideal fusion method to get the final classification. In the experiments, 5‐fold cross‐validation method is used to validate the proposed method on 1,600 pulmonary nodule images and our method reaches high‐classification sensitivities of 91.75% at 1.58 false positives per scan. Meanwhile, compared with other deep learning diagnosis methods, our method achieves better discriminative results and is highly suited to be used for pulmonary nodule diagnosis.

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

“…The advantage of MTANN is fewer training cases compared to CNN without compromising classification performance [20]. SDAE-ELM is a feature vector deep learning algorithm combined with ELM, which is a feed-forward neural network [37]. The advantages of stacked autoencoders include fewer training cases compared to, for example, CNN, since stacked autoencoders are able to generate new images from the image characteristic feature vectors [44].…”

“…[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)…”

“…For studies that only reported results on classification performance, five studies [34,35,37,40,42] tested on local, independently obtained datasets. All studies provided reports of accuracy, which ranged between 68-92%.…”

“…All studies provided reports of accuracy, which ranged between 68-92%. Four of these studies [34,35,40,42] had deep learning architectures based on CNN, while only Qiang et al [37] used SDAE-ELM. For Nishio et al [34], sensitivity and specificity were calculated from values given in a confusion matrix for benign, primary cancer, and metastatic cancer as 50.1% and 84.4%, 77.6% and 77.4%, and 74% and 88.2%, respectively.…”

A multiperiod hybrid decision support model for medical diagnosis and treatment based on similarities and three‐way decision theory

“…Medical diagnosis is considered to be a convenient support tool in clinical medicine that helps physicians to determine the most possible disease and give appropriate medicated figures on the basis of a set of given symptoms. In the last few years, numerous approaches have been introduced to address medical diagnosis problems in an efficient way, including learning machine (Gürbüz & Kılıç, 2014;Qiang, Ge, Zhao, Zhang, & Tang, 2017), case-based reasoning (Chattopadhyay, Banerjee, Rabhi, & Acharya, 2013;Park, Kim, & Chun, 2006), Bayesian network (Zarikas, Papageorgiou, & Regner, 2015), and statistical or pattern recognition methods (Hemanth, Anitha, & Ane, 2017;Wolfers, Buitelaar, Beckmann, Franke, & Marquand, 2015), among others. Nonetheless, a crucial issue in medical practice is that patients' original information is usually imprecise and uncertain because the collection of information is expensive, time-consuming, and even harmful to patients.…”