Abstract
Background: Accurate prognostic stratification as well as segmentation of Head-and-Neck Squamous-Cell-Carcinoma (HNSCC) patients can be an important clinical reference when designing therapeutic strategies. We set to enable automated segmentation of tumors and prediction of recurrence-free survival (RFS) using advanced deep learning techniques and Hybrid Machine Learning Systems (HMLSs).
Method: In this work, 883 subjects were extracted from HECKTOR-Challenge: ~60% of the total subjects were considered for the training and validation procedure, and the remaining subjects for external testing were employed to validate our models. PET images were registered to CT. First, a weighted fusion technique was employed to combine PET and CT information. We also employed Cascade-Net to enable automated segmentation of HNSCC tumors. Moreover, we extracted deep learning features (DF) via a 3D auto-encoder algorithm from PET and the fused image. Subsequently, we employed an HMLS including a feature selection algorithm such as Mutual Information (MI) linked with a survival prediction algorithm such as Random Survival Forest (RSF) optimized by 5-fold cross-validation and grid search. The dataset with DFs was normalized by the z-score technique. Moreover, dice score and c-Index were reported to evaluate the segmentation and prediction models, respectively.
Result: For segmentation, the weighted fusion technique followed by the Cascade-Net segmentation algorithm resulted in a validation dice score of 72%. External testing of 71% confirmed our findings. DFs extracted from sole PET and MI followed by RSF enabled us to receive a validation c-index of 66% for RFS prediction. The external testing of 59% confirmed our finding.
Conclusion: We demonstrated that using the fusion technique followed by an appropriate automated segmentation technique provides good performance. Moreover, employing DFs extracted from sole PET and HMLS, including MI linked with RSF, enables us to perform the appropriate survival prediction. We also showed imaging information extracted from PET outperformed the usage of the fused images in the prediction of RFS.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Fatan, M., Hosseinzadeh, M., Askari, D., Sheikhi, H., Rezaeijo, S.M., Salmanpour, M.R.: Fusion-based head and neck tumor segmentation and survival prediction using robust deep learning techniques and advanced hybrid machine learning systems. In: Andrearczyk, V., Oreiller, V., Hatt, M., Depeursinge, A. (eds.) HECKTOR 2021. LNCS, vol. 13209, pp. 211–223. Springer, Cham (2022). https://doi.org/10.1007/978-3-030-98253-9_20
Salmanpour, M.R., Hajianfar, G., Rezaeijo, S.M., Ghaemi, M., Rahmim, A.: Advanced automatic segmentation of tumors and survival prediction in head and neck cancer. In: Andrearczyk, V., Oreiller, V., Hatt, M., Depeursinge, A. (eds.) HECKTOR 2021. LNCS, vol. 13209, pp. 202–210. Springer, Cham (2022). https://doi.org/10.1007/978-3-030-98253-9_19
Singh, Y., Bhatia, P.K., Sangwan, O.: A review of studies on machine learning techniques. Int. J. Comput. Sci. Secur. 1(1), 70–84 (2007)
Salmanpour, M.R., Shamsaei, M., Rahmim, A.: Feature selection and machine learning methods for optimal identification and prediction of subtypes in Parkinson’s disease. Comput. Methods Programs Biomed. 206, 106131 (2021)
Salmanpour, M.R., et al.: Optimized machine learning methods for prediction of cognitive outcome in Parkinson’s disease. Comput. Biol. Med. 111, 103347 (2019)
Salmanpour, M.R., et al.: Machine learning methods for optimal prediction of motor outcome in Parkinson’s disease. Physica Med. 69, 233–240 (2020)
Oreiller, V., et al.: Head and neck tumor segmentation in PET/CT: the HECKTOR challenge. Med. Image Anal. 77, 102336 (2022)
Iddi, S., Li, D., Aisen, P.S., Rafii, M.S., Thompson, W.K., Donohue, M.C.: Predicting the course of Alzheimer’s progression. Brain Inform. 6(1), 1–18 (2019). https://doi.org/10.1186/s40708-019-0099-0
Vallieres, M., et al.: Radiomics strategies for risk assessment of tumour failure in head-and-neck cancer. Sci. Rep. 7(1), 1–14 (2017)
Javanmardi, A., Hosseinzadeh, M., Hajianfar, G., Nabizadeh, A.H., Rezaeijo, S.M., Rahmim, A., Salmanpour, M.: Multi-modality fusion coupled with deep learning for improved outcome prediction in head and neck cancer. In: Book Multi-Modality Fusion Coupled with Deep Learning for Improved Outcome Prediction in Head and Neck Cancer, pp. 664–668. SPIE (2022)
Butowski, N.A.: Epidemiology and diagnosis of brain tumors. CONTINUUM: Lifelong Learn. Neurol. 21(2), 301–313 (2015)
Kumari, N., Saxena, S.: Review of brain tumor segmentation and classification. In: Book Review of Brain Tumor Segmentation and Classification, pp. 1–6. IEEE (2018)
Daisne, J.-F., et al.: Tumor volume in pharyngolaryngeal squamous cell carcinoma: comparison at CT, MR imaging, and FDG PET and validation with surgical specimen. Radiology 233(1), 93–100 (2004)
Rodrigues, R.S., et al.: Comparison of whole-body PET/CT, dedicated high-resolution head and neck PET/CT, and contrast-enhanced CT in preoperative staging of clinically M0 squamous cell carcinoma of the head and neck. J. Nucl. Med. 50(8), 1205–1213 (2009)
Roh, J.-L., et al.: 2-[18F]-Fluoro-2-deoxy-D-glucose positron emission tomography as guidance for primary treatment in patients with advanced-stage resectable squamous cell carcinoma of the larynx and hypopharynx. Eur. J. Surg. Oncol. (EJSO) 33(6), 790–795 (2007)
Eyassu, E., Young, M.: Nuclear medicine PET/CT head and neck cancer assessment, protocols, and interpretation. StatPearls [Internet] (2022). StatPearls Publishing
Taxak, N., Scholar, M.T., Singhal, S.: A Review of Image Fusion Methods
Wang, Q., Shen, Y., Jin, J.: Performance evaluation of image fusion techniques. Image Fusion: Algorithms Appl. 19, 469–492 (2008)
Salmanpour, M.R., Hajianfar, G., Lv, W., Lu, L., Rahmim, A.: Multitask outcome prediction using hybrid machine learning and PET-CT fusion radiomics. In: Book Multitask Outcome Prediction using Hybrid Machine Learning and PET-CT Fusion Radiomics (2021). Soc. Nuclear. Med.
Zwanenburg, A., et al.: The image biomarker standardization initiative: standardized quantitative radiomics for high-throughput image-based phenotyping. Radiology 295(2), 328–338 (2020)
Salmanpour, M.R., Shamsaei, M., Saberi, A., Hajianfar, G., Soltanian-Zadeh, H., Rahmim, A.: Robust identification of Parkinson’s disease subtypes using radiomics and hybrid machine learning. Comput. Biol. Med. 129, 104142 (2021)
Bank, D., Koenigstein, N., Giryes, R.: Autoencoders. arXiv preprint arXiv:2003.05991 (2020)
Rahmim, A., Zaidi, H.: PET versus SPECT: strengths, limitations and challenges. Nucl. Med. Commun. 29(3), 193–207 (2008)
Wang, G., Huang, Z., Shen, H., Hu, Z.: The head and neck tumor segmentation in PET/CT based on multi-channel attention network. In: Andrearczyk, V., Oreiller, V., Hatt, M., Depeursinge, A. (eds.) Head and Neck Tumor Segmentation and Outcome Prediction. HECKTOR 2021. Lecture Notes in Computer Science, vol. 13209, pp. 68–74. Springer, Cham (2022). https://doi.org/10.1007/978-3-030-98253-9_5
Xie, J., Peng, Y.: The head and neck tumor segmentation based on 3D U-Net. In: Andrearczyk, V., Oreiller, V., Hatt, M., Depeursinge, A. (eds.) HECKTOR 2021. LNCS, vol. 13209, pp. 92–98. Springer, Cham (2022). https://doi.org/10.1007/978-3-030-98253-9_8
Salmanpour, M.R., et al.: Advanced survival prediction in head and neck cancer using hybrid machine learning systems and radiomics features. In: Book Advanced Survival Prediction in Head and Neck Cancer Using Hybrid Machine Learning Systems and Radiomics Features, pp. 314–321. SPIE (2022)
Salmanpour, M.R., et al.: Prediction of TNM stage in head and neck cancer using hybrid machine learning systems and radiomics features. In: Book Prediction of TNM Stage in Head and Neck Cancer Using Hybrid Machine Learning Systems and Radiomics Features, pp. 648–653. SPIE (2022)
Salmanpour, M.R., et al.: Deep versus handcrafted tensor radiomics features: application to survival prediction in head and neck cancer. In: EANM (2022)
Tang, M., Zhang, Z., Cobzas, D., Jagersand, M., Jaremko, J.L.: Segmentation-by-detection: a cascade network for volumetric medical image segmentation. In: Book Segmentation-By-Detection: A Cascade Network for Volumetric Medical Image Segmentation, pp. 1356–1359. IEEE (2018)
De Jay, N., Papillon, S., Olsen, C., El-, N., Bontempi, G., Haibe-Kains, B.: mRMRe: an R package for parallelized mRMR ensemble feature selection. Bioinformatics 29(18), 2365–2368 (2013)
Ishwaran, H., Kogalur, U.B., Blackstone, E.H., Lauer, M.S.: Random survival forests. Ann. Appl. Stat. 2(3), 841–860 (2008)
Acknowledgment
This study was supported by the Technological Virtual Collaboration Corporation (TECVICO Corp.), Vancouver, BC, Canada, as well as the Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery Grant RGPIN-2019–06467.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Ethics declarations
Code Availability
All codes are publicly shared at: https://github.com/Tecvico.
Conflict of Interest
The authors have no relevant conflicts of interest to disclose.
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Salmanpour, M.R. et al. (2023). Deep Learning and Machine Learning Techniques for Automated PET/CT Segmentation and Survival Prediction in Head and Neck Cancer. In: Andrearczyk, V., Oreiller, V., Hatt, M., Depeursinge, A. (eds) Head and Neck Tumor Segmentation and Outcome Prediction. HECKTOR 2022. Lecture Notes in Computer Science, vol 13626. Springer, Cham. https://doi.org/10.1007/978-3-031-27420-6_23
Download citation
DOI: https://doi.org/10.1007/978-3-031-27420-6_23
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-27419-0
Online ISBN: 978-3-031-27420-6
eBook Packages: Computer ScienceComputer Science (R0)
