Lung cancer is the leading cause of cancer deaths worldwide. Accurate early diagnosis is critical
in increasing the 5-year survival rate of lung cancer, so the efficient and accurate detection of lung nodules,
the potential precursors to lung cancer, is paramount. In this paper, a computer-aided lung nodule detection
system using 3D deep convolutional neural networks (CNNs) is developed. The first multi-scale 11-layer 3D fully
convolutional neural network (FCN) is used for screening all lung nodule candidates. Considering relative small
sizes of lung nodules and limited memory, the input of the FCN consists of 3D image patches rather than of whole
images. The candidates are further classified in the second CNN to get the final result. The proposed method
achieves high performance in the LUNA16 challenge and demonstrates the effectiveness of using 3D deep CNNs
for lung nodule detection.
FU Ling (傅玲), MA Jingchen (马璟琛), CHEN Yizhi (琛奕志), LARSSON Rasmus, ZHAO Jun *(赵俊)
. Automatic Detection of Lung Nodules Using 3D Deep Convolutional Neural Networks[J]. Journal of Shanghai Jiaotong University(Science), 2019
, 24(4)
: 517
-523
.
DOI: 10.1007/s12204-019-2084-4
[1] SIEGEL R L, MILLER K D, JEMAL A. Cancer statistics,2016 [J]. CA: A Cancer Journal for Clinicians,2016, 66(1): 7-30.
[2] WANG B, TIAN X D, WANG Q, et al. Pulmonarynodule detection in CT images based on shape constraintCV model [J]. Medical Physics, 2015, 42(3):1241-1254.
[3] LI Q, SONE S, DOI K. Selective enhancement filtersfor nodules, vessels, and airway walls in two- andthree-dimensional CT scans [J]. Medical Physics, 2003,30(8): 2040-2051.
[4] TAN M, DEKLERCK R, JANSEN B, et al. A novelcomputer-aided lung nodule detection system for CTimages [J]. Medical Physics, 2011, 38(10): 5630-5645.
[5] MESSAY T, HARDIE R C, ROGERS S K. A new computationallyefficient CAD system for pulmonary noduledetection in CT imagery [J]. Medical Image Analysis,2010, 14(3): 390-406.
[6] SETIO A A A, CIOMPI F, LITJENS G, et al. Pulmonarynodule detection in CT images: False positivereduction using multi-view convolutional networks [J].IEEE Transactions on Medical Imaging, 2016, 35(5):1160-1169.
[7] MURPHY K, VAN GINNEKEN B, SCHILHAM A MR, et al. A large-scale evaluation of automatic pulmonarynodule detection in chest CT using local imagefeatures and k-nearest-neighbour classification [J].Medical Image Analysis, 2009, 13(5): 757-770.
[8] JACOBS C, VAN RIKXOORT E M, TWELLMANNT, et al. Automatic detection of subsolid pulmonarynodules in thoracic computed tomography images [J].Medical Image Analysis, 2014, 18(2): 374-384.
[9] SETIO A A A, JACOBS C, GELDERBLOM J, etal. Automatic detection of large pulmonary solid nodulesin thoracic CT images [J]. Medical Physics, 2015,42(10): 5642-5653.
[10] CIOMPI F, JACOBS C, SCHOLTEN E T, et al. Bagof-frequencies: A descriptor of pulmonary nodules incomputed tomography images [J]. IEEE Transactionson Medical Imaging, 2015, 34(4): 962-973.
