Deep learning approach for cancer subtype classification using high-dimensional gene expression data

Abstract

Motivation Studies have shown that classifying cancer subtypes can provide valuable information for a range of cancer research, from aetiology and tumour biology to prognosis and personalized treatment. Current methods usually adopt gene expression data to perform cancer subtype classification. However, cancer samples are scarce, and the high-dimensional features of their gene expression data are too sparse to allow most methods to achieve desirable classification results. Results In this paper, we propose a deep learning approach by combining a convolutional neural network (CNN) and bidirectional gated recurrent unit (BiGRU): our approach, DCGN, aims to achieve nonlinear dimensionality reduction and learn features to eliminate irrelevant factors in gene expression data. Specifically, DCGN first uses the synthetic minority oversampling technique algorithm to equalize data. The CNN can handle high-dimensional data without stress and extract important local features, and the BiGRU can analyse deep features and retain their important information; the DCGN captures key features by combining both neural networks to overcome the challenges of small sample sizes and sparse, high-dimensional features. In the experiments, we compared the DCGN to seven other cancer subtype classification methods using breast and bladder cancer gene expression datasets. The experimental results show that the DCGN performs better than the other seven methods and can provide more satisfactory classification results.

Deep learning approach for brain tumor classification using metaheuristic optimization with gene expression data

“…We compare the performance of the proposed system to a number of different current schemes of classification using gene expression datasets in this section. We specifically compared our proposed approach to those put forth by Alrefai et al, 45 Majji et al, 46 Balcı et al, 47 Shen et al, 48 and Wang et al, 49 Almarzouki et al, 50 Aziz et al, 8 Hanczar et al 51 We measured each scheme's performance at 86.7, 91.76, 90.07, 86.62, 94.32, 90.24, and 89.23, respectively, while our proposed model performance is 98.70. These findings are summarized Table 4.…”

Applications of artificial intelligence in urologic oncology

“…Each study’s inclusion and exclusion criteria were highly heterogeneous, but they all focused on urological cancer. Finally, 58 studies on urological cancers (prostate cancer: 21 [ 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 ], bladder cancer: 20 [ 5 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 ], kidney cancer: 17 [ 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 ]) were identified as shown in Fig. 1 .…”

Improving Cancer Classification Using Deep Reinforcement Learning with Convolutional LSTM Networks

“…An effective DL model called DCGN was presented [28] by integrating CNN and bidirectional gated recurrent unit (BiGRU) for cancer subtypes classification. BiGRU analyzes deep features, cancer subtype classification is improved by DCGN's ability to deal with high-dimensional data and extract local characteristics.…”

“…Additionally, a comparative analysis is conducted between proposed and earlier models implemented on the considered literature: 1D-CNN [26], DL-DCGN [28] DL-SAE [29] and DL-AAA [31] regarding the following metrics:…”