Deep learning vs. atlas-based models for fast auto-segmentation of the masticatory muscles on head and neck CT images

被引:45
作者
Chen, Wen [1 ,2 ]
Li, Yimin [3 ]
Dyer, Brandon A. [2 ,4 ]
Feng, Xue [5 ]
Rao, Shyam [2 ]
Benedict, Stanley H. [2 ]
Chen, Quan [5 ,6 ]
Rong, Yi [2 ]
机构
[1] Cent South Univ, Xiangya Hosp, Dept Radiat Oncol, Changsha, Peoples R China
[2] Univ Calif Davis, Dept Radiat Oncol, Med Ctr, 4501 X St,Suite 0152, Sacramento, CA 95817 USA
[3] Xiamen Univ, Affiliated Hosp 1, Dept Radiat Oncol, Xiamen Canc Ctr, Xiamen, Fujian, Peoples R China
[4] Univ Washington, Dept Radiat Oncol, Seattle, WA 98195 USA
[5] Carina Med LLC, 145 Graham Ave,A168, Lexington, KY 40536 USA
[6] Univ Kentucky, Markey Canc Ctr, Dept Radiat Oncol, RM CC063,800 Rose St, Lexington, KY 40536 USA
关键词
Deep learning model; Masticatory muscles; Auto-segmentation; RADIATION-INDUCED TRISMUS; CLINICAL TARGET VOLUME; QUALITY-OF-LIFE; INTEROBSERVER VARIABILITY; CANCER; DELINEATION; ORGANS; RISK; ALGORITHM; SOFTWARE;
D O I
10.1186/s13014-020-01617-0
中图分类号
R73 [肿瘤学];
学科分类号
100214 ;
摘要
Background Impaired function of masticatory muscles will lead to trismus. Routine delineation of these muscles during planning may improve dose tracking and facilitate dose reduction resulting in decreased radiation-related trismus. This study aimed to compare a deep learning model with a commercial atlas-based model for fast auto-segmentation of the masticatory muscles on head and neck computed tomography (CT) images. Material and methods Paired masseter (M), temporalis (T), medial and lateral pterygoid (MP, LP) muscles were manually segmented on 56 CT images. CT images were randomly divided into training (n = 27) and validation (n = 29) cohorts. Two methods were used for automatic delineation of masticatory muscles (MMs): Deep learning auto-segmentation (DLAS) and atlas-based auto-segmentation (ABAS). The automatic algorithms were evaluated using Dice similarity coefficient (DSC), recall, precision, Hausdorff distance (HD), HD95, and mean surface distance (MSD). A consolidated score was calculated by normalizing the metrics against interobserver variability and averaging over all patients. Differences in dose ( increment Dose) to MMs for DLAS and ABAS segmentations were assessed. A paired t-test was used to compare the geometric and dosimetric difference between DLAS and ABAS methods. Results DLAS outperformed ABAS in delineating all MMs (p < 0.05). The DLAS mean DSC for M, T, MP, and LP ranged from 0.83 +/- 0.03 to 0.89 +/- 0.02, the ABAS mean DSC ranged from 0.79 +/- 0.05 to 0.85 +/- 0.04. The mean value for recall, HD, HD95, MSD also improved with DLAS for auto-segmentation. Interobserver variation revealed the highest variability in DSC and MSD for both T and MP, and the highest scores were achieved for T by both automatic algorithms. With few exceptions, the mean increment D98%, increment D95%, increment D50%, and increment D2% for all structures were below 10% for DLAS and ABAS and had no detectable statistical difference (P > 0.05). DLAS based contours had dose endpoints more closely matched with that of the manually segmented when compared with ABAS. Conclusions DLAS auto-segmentation of masticatory muscles for the head and neck radiotherapy had improved segmentation accuracy compared with ABAS with no qualitative difference in dosimetric endpoints compared to manually segmented contours.
引用
收藏
页数:10
相关论文
共 39 条
[1]   3D Variation in delineation of head and neck organs at risk [J].
Brouwer, Charlotte L. ;
Steenbakkers, Roel J. H. M. ;
van den Heuvel, Edwin ;
Duppen, Joop C. ;
Navran, Arash ;
Bijl, Henk P. ;
Chouvalova, Olga ;
Burlage, Fred R. ;
Meertens, Harm ;
Langendijk, Johannes A. ;
van 't Veld, Aart A. .
RADIATION ONCOLOGY, 2012, 7
[2]   Head and neck cancer patient images for determining auto-segmentation accuracy in T2-weighted magnetic resonance imaging through expert manual segmentations [J].
Cardenas, Carlos E. ;
Mohamed, Abdallah S. R. ;
Yang, Jinzhong ;
Gooding, Mark ;
Veeraraghavan, Harini ;
Kalpathy-Cramer, Jayashree ;
Ng, Sweet Ping ;
Ding, Yao ;
Wang, Jihong ;
Lai, Stephen Y. ;
Fuller, Clifton D. ;
Sharp, Greg .
MEDICAL PHYSICS, 2020, 47 (05) :2317-2322
[3]  
Cicek Ozgun, 2016, Medical Image Computing and Computer-Assisted Intervention - MICCAI 2016. 19th International Conference. Proceedings: LNCS 9901, P424, DOI 10.1007/978-3-319-46723-8_49
[4]   Atlas-based automatic segmentation of head and neck organs at risk and nodal target volumes: a clinical validation [J].
Daisne, Jean-Francois ;
Blumhofer, Andreas .
RADIATION ONCOLOGY, 2013, 8
[5]   Assessment of fully-automated atlas-based segmentation of novel oral mucosal surface organ-at-risk [J].
Dean, Jamie A. ;
Welsh, Liam C. ;
McQuaid, Dualta ;
Wong, Kee H. ;
Aleksic, Aleksandar ;
Dunne, Emma ;
Islam, Mohammad R. ;
Patel, Anushka ;
Patel, Priyanka ;
Petkar, Imran ;
Phillips, Iain ;
Sham, Jackie ;
Newbold, Kate L. ;
Bhide, Shreerang A. ;
Harrington, Kevin J. ;
Gulliford, Sarah L. ;
Nutting, Christopher M. .
RADIOTHERAPY AND ONCOLOGY, 2016, 119 (01) :166-171
[6]   Comparison of Automated Atlas-Based Segmentation Software for Postoperative Prostate Cancer Radiotherapy [J].
Delpon, Gregory ;
Escande, Alexandre ;
Ruef, Timothee ;
Darreon, Julien ;
Fontaine, Jimmy ;
Noblet, Caroline ;
Supiot, Stephane ;
Lacornerie, Thomas ;
Pasquier, David .
FRONTIERS IN ONCOLOGY, 2016, 6
[7]   Validation of clinical acceptability of an atlas-based segmentation algorithm for the delineation of organs at risk in head and neck cancer [J].
Duc, Albert K. Hoang ;
Eminowicz, Gemma ;
Mendes, Ruheena ;
Wong, Swee-Ling ;
McClelland, Jamie ;
Modat, Marc ;
Cardoso, M. Jorge ;
Mendelson, Alex F. ;
Veiga, Catarina ;
Kadir, Timor ;
D'Souza, Derek ;
Ourselin, Sebastien .
MEDICAL PHYSICS, 2015, 42 (09) :5027-5034
[8]  
Feng X., 2019, MED PHYS
[9]   Improving accuracy and robustness of deep convolutional neural network based thoracic OAR segmentation [J].
Feng, Xue ;
Bernard, Mark E. ;
Hunter, Thomas ;
Chen, Quan .
PHYSICS IN MEDICINE AND BIOLOGY, 2020, 65 (07)
[10]   Intra- and inter-observer variability in contouring prostate and seminal vesicles: implications for conformal treatment planning [J].
Fiorino, C ;
Reni, M ;
Bolognesi, A ;
Cattaneo, GM ;
Calandrino, R .
RADIOTHERAPY AND ONCOLOGY, 1998, 47 (03) :285-292