TY - JOUR
T1 - Automatic multi-organ segmentation in computed tomography images using hierarchical convolutional neural network
AU - Sultana, Sharmin
AU - Robinson, Adam
AU - Song, Daniel Y.
AU - Lee, Junghoon
N1 - Funding Information:
This work was supported by the National Cancer Institute (NCI), National Institutes of Health (NIH) under Grant No. R01CA151395.
Publisher Copyright:
© 2020 Society of Photo-Optical Instrumentation Engineers (SPIE)
PY - 2020/9/1
Y1 - 2020/9/1
N2 - Purpose: Accurate segmentation of treatment planning computed tomography (CT) images is important for radiation therapy (RT) planning. However, low soft tissue contrast in CT makes the segmentation task challenging. We propose a two-step hierarchical convolutional neural network (CNN) segmentation strategy to automatically segment multiple organs from CT. Approach: The first step generates a coarse segmentation from which organ-specific regions of interest (ROIs) are produced. The second step produces detailed segmentation of each organ. The ROIs are generated using UNet, which automatically identifies the area of each organ and improves computational efficiency by eliminating irrelevant background information. For the fine segmentation step, we combined UNet with a generative adversarial network. The generator is designed as a UNet that is trained to segment organ structures and the discriminator is a fully convolutional network, which distinguishes whether the segmentation is real or generator-predicted, thus improving the segmentation accuracy. We validated the proposed method on male pelvic and head and neck (H&N) CTs used for RT planning of prostate and H&N cancer, respectively. For the pelvic structure segmentation, the network was trained to segment the prostate, bladder, and rectum. For H&N, the network was trained to segment the parotid glands (PG) and submandibular glands (SMG). Results: The trained segmentation networks were tested on 15 pelvic and 20 H&N independent datasets. The H&N segmentation network was also tested on a public domain dataset (N = 38) and showed similar performance. The average dice similarity coefficients (mean ± SD) of pelvic structures are 0.91 ± 0.05 (prostate), 0.95 ± 0.06 (bladder), 0.90 ± 0.09 (rectum), and H&N structures are 0.87 ± 0.04 (PG) and 0.86 ± 0.05 (SMG). The segmentation for each CT takes <10 s on average. Conclusions: Experimental results demonstrate that the proposed method can produce fast, accurate, and reproducible segmentation of multiple organs of different sizes and shapes and show its potential to be applicable to different disease sites.
AB - Purpose: Accurate segmentation of treatment planning computed tomography (CT) images is important for radiation therapy (RT) planning. However, low soft tissue contrast in CT makes the segmentation task challenging. We propose a two-step hierarchical convolutional neural network (CNN) segmentation strategy to automatically segment multiple organs from CT. Approach: The first step generates a coarse segmentation from which organ-specific regions of interest (ROIs) are produced. The second step produces detailed segmentation of each organ. The ROIs are generated using UNet, which automatically identifies the area of each organ and improves computational efficiency by eliminating irrelevant background information. For the fine segmentation step, we combined UNet with a generative adversarial network. The generator is designed as a UNet that is trained to segment organ structures and the discriminator is a fully convolutional network, which distinguishes whether the segmentation is real or generator-predicted, thus improving the segmentation accuracy. We validated the proposed method on male pelvic and head and neck (H&N) CTs used for RT planning of prostate and H&N cancer, respectively. For the pelvic structure segmentation, the network was trained to segment the prostate, bladder, and rectum. For H&N, the network was trained to segment the parotid glands (PG) and submandibular glands (SMG). Results: The trained segmentation networks were tested on 15 pelvic and 20 H&N independent datasets. The H&N segmentation network was also tested on a public domain dataset (N = 38) and showed similar performance. The average dice similarity coefficients (mean ± SD) of pelvic structures are 0.91 ± 0.05 (prostate), 0.95 ± 0.06 (bladder), 0.90 ± 0.09 (rectum), and H&N structures are 0.87 ± 0.04 (PG) and 0.86 ± 0.05 (SMG). The segmentation for each CT takes <10 s on average. Conclusions: Experimental results demonstrate that the proposed method can produce fast, accurate, and reproducible segmentation of multiple organs of different sizes and shapes and show its potential to be applicable to different disease sites.
KW - Deep learning
KW - Hierarchical convolutional neural network
KW - Radiotherapy
KW - Segmentation
UR - http://www.scopus.com/inward/record.url?scp=85096552761&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85096552761&partnerID=8YFLogxK
U2 - 10.1117/1.JMI.7.5.055001
DO - 10.1117/1.JMI.7.5.055001
M3 - Article
C2 - 33102622
AN - SCOPUS:85096552761
SN - 2329-4302
VL - 7
JO - Journal of Medical Imaging
JF - Journal of Medical Imaging
IS - 5
M1 - 055001
ER -