A Novel Framework with High Diagnostic Sensitivity for Lung Cancer Detection by Electronic Nose

Abstract

The electronic nose (e-nose) system is a newly developing detection technology for its advantages of non-invasiveness, simple operation, and low cost. However, lung cancer screening through e-nose requires effective pattern recognition frameworks. Existing frameworks rely heavily on hand-crafted features and have relatively low diagnostic sensitivity. To handle these problems, gated recurrent unit based autoencoder (GRU-AE) is adopted to automatically extract features from temporal and high-dimensional e-nose data. Moreover, we propose a novel margin and sensitivity based ordering ensemble pruning (MSEP) model for effective classification. The proposed heuristic model aims to reduce missed diagnosis rate of lung cancer patients while maintaining a high rate of overall identification. In the experiments, five state-of-the-art classification models and two popular dimensionality reduction methods were involved for comparison to demonstrate the validity of the proposed GRU-AE-MSEP framework, through 214 collected breath samples measured by e-nose. Experimental results indicated that the proposed intelligent framework achieved high sensitivity of 94.22%, accuracy of 93.55%, and specificity of 92.80%, thereby providing a new practical means for wide disease screening by e-nose in medical scenarios.

Hybrid Network with Attention Mechanism for Detection and Location of Myocardial Infarction Based on 12-Lead Electrocardiogram Signals

“…Sensitivity (Sen) measures the proportion of real MI patients who are correctly classified, and defined as Equation (15). Instead, specificity (Spe), defined in Equation 16…”

“…In recent decades, deep learning methods, including convolutional neural network (CNN), gated recurrent unit (GRU), attention mechanism, and autoencoder, have been widely and superbly applied to analyze biomedical signals [14][15][16]. Instead of separate feature extraction and classification processes, deep learning architectures automatically extract critical features required for classification from vast samples [17].…”

Recent Progress in Smart Electronic Nose Technologies Enabled with Machine Learning Methods

“…After the model was trained, an extra prediction layer was appended to the encoder of autoencoder for prediction. Lu et al [63] replaced the hand-craft features for the E-Nose with latent representation generated from a gated recurrent unit-based autoencoder (GRU-AE). Compared to other dimensionality reduction methods including PCA and Kernel-PCA, feature representations from the GRU-AE were more distinguishable and effectively improved classification performance.…”

A Review of Exhaled Volatile Organic Compounds as Biomarkers for Thoracic Malignancies

“…A diagnostic model by Oguma et al [59] was able to determine lung cancer with 75% sensitivity and 78% specificity, which decreased slightly to 73% sensitivity and 78% specificity for early lung cancer. Lu et al [54] analyzed 214 breath samples by e-nose and was able to correctly identify LC with 94.2% sensitivity and 92.8% specificity. The authors detected stage II with 97.9% sensitivity and 70% specificity, stage III with 82.8% sensitivity and 81.8% specificity, and stage IV with 83.2% sensitivity and 81.6% specificity.…”