TY - GEN
T1 - Lung nodule classification based on deep learning networks and handcraft segmentation
AU - Salvador-Torres, Luis G.
AU - Almaraz-Damian, Jose A.
AU - Ponomaryov, Volodymyr I.
AU - Reyes-Reyes, Rogelio
AU - Cruz-Ramos, Clara
N1 - Publisher Copyright:
© 2022 SPIE.
PY - 2022
Y1 - 2022
N2 - This study proposes a hybrid CAD system, where the first stage consists of the handcraft segmentation, following a CNN based on the ResNet-34 architecture. In the segmentation stage, the rib cage (thorax region) is extracted using the K-means algorithm. The extraction of the nodules is performed in two steps, those attached to the pleura are found via a K-means clustering on the rib cage, and the circumscribed and vascular nodules are extracted using morphological operations. The resulting segmentation masks are applied to the test images, decreasing the number of false positives. Finally, the resulting image is splitted of the patches to be classified by the ResNet-34 trained from scratch. Designed CAD system has been implemented on Google Colab platform and a standalone computer with Nvidia®RTX 3090. The experiments with different CAD systems were performed on SPIE and LIDC-IDRI datasets demonstrating better performance of designed technique with reduction of false-positive objects.
AB - This study proposes a hybrid CAD system, where the first stage consists of the handcraft segmentation, following a CNN based on the ResNet-34 architecture. In the segmentation stage, the rib cage (thorax region) is extracted using the K-means algorithm. The extraction of the nodules is performed in two steps, those attached to the pleura are found via a K-means clustering on the rib cage, and the circumscribed and vascular nodules are extracted using morphological operations. The resulting segmentation masks are applied to the test images, decreasing the number of false positives. Finally, the resulting image is splitted of the patches to be classified by the ResNet-34 trained from scratch. Designed CAD system has been implemented on Google Colab platform and a standalone computer with Nvidia®RTX 3090. The experiments with different CAD systems were performed on SPIE and LIDC-IDRI datasets demonstrating better performance of designed technique with reduction of false-positive objects.
KW - CNN
KW - Computer-Aided Detection
KW - LIDC-IDRI dataset
KW - Lung cancer
KW - SPIE dataset
KW - computed tomography
UR - http://www.scopus.com/inward/record.url?scp=85135709179&partnerID=8YFLogxK
U2 - 10.1117/12.2618176
DO - 10.1117/12.2618176
M3 - Contribución a la conferencia
AN - SCOPUS:85135709179
T3 - Proceedings of SPIE - The International Society for Optical Engineering
BT - Real-Time Image Processing and Deep Learning 2022
A2 - Kehtarnavaz, Nasser
A2 - Carlsohn, Matthias F.
PB - SPIE
T2 - Real-Time Image Processing and Deep Learning 2022
Y2 - 6 June 2022 through 12 June 2022
ER -