Classification Method of Benign and Malignant Pulmonary Nodules Based on MDRA-net

Manman Fei; Chunxiao Chen; Liang Wang; Xue Fu

doi:10.3788/LOP212753

Laser & Optoelectronics Progress, Volume. 60, Issue 4, 0417001(2023)

Classification Method of Benign and Malignant Pulmonary Nodules Based on MDRA-net

Manman Fei, Chunxiao Chen^*, Liang Wang, and Xue Fu

Department of Biomedical Engineering, College of Automation Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing 211106, Jiangsu, China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(6)

Fig. 1. Structure diagram of three-dimensional multi-depth-residual attention networks

Download full size

Fig. 2. Structure diagram of PE block

Download full size

Fig. 3. Structure diagram of multi-depth fusion block

Download full size

Fig. 4. Some examples of misjudgment of malignant nodules in different classification models. (a1)-(a5) 3D CNN misjudged, MDR-net and MDRA judged correctly; (b1)-(b5) 3D CNN and MDR-net misjudged, and MDRA judged correctly; (c1)-(c5) 3D CNN, MDR-net, and MDRA-net all misjudged

Download full size

Table 1. Comparison of experimental results
View table
Table 1. Comparison of experimental results
Model Accuracy Sensitivity Specificity
3D CNN 96.34 87.41 99.50
MDR-net 95.79 90.90 97.52
MDRA-net 96.52 93.01 97.77

Table 2. Comparison of different classification methods of pulmonary nodules

View table

Table 2. Comparison of different classification methods of pulmonary nodules

Method	Accuracy /%	Sensitivity /%	Specificity /%	AUC /10^-2
Hua et al^［6］		82.20	73.40
Han et al^［17］	85.38	70.20	92.80	88.19
Shen et al^［8］	87.14	77.00	93.00
Xie et al^［18］	92.53	84.94	96.28	95.81
MoDenseNet^［10］	90.40	90.47	90.33	95.48
DenseNet_centercrop^［11］	87.48			92.71
SE-CapsNet^［12］	95.83	92.10	93.50
method Proposed	96.52	93.01	97.77	98.11

Tools

Get Citation

Copy Citation Text

Manman Fei, Chunxiao Chen, Liang Wang, Xue Fu. Classification Method of Benign and Malignant Pulmonary Nodules Based on MDRA-net[J]. Laser & Optoelectronics Progress, 2023, 60(4): 0417001

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites