Prognostic Analysis Model of Renal Clear Cell Carcinoma Based on Multi-Dictionary Learning
Tu Chao1,2, Ning Zhenyuan1,2, Zhang Yu1,2*
1(School of Biomedical Engineering, Southern Medical University, Guangzhou 510515, China) 2(Guangdong Provincial Key Laboratory of Medical Image Processing, Southern Medical University, Guangzhou 510515, China)
Abstract：Clear cell renal cell carcinoma (ccRCC) is a highly heterogeneous tumor with complex and variable clinical manifestations. Automatic histopathological whole slide image (WSI) analysis is a useful approach for pathologists to make diagnosis. However, feature extraction for the prognostic analysis of ccRCC is a challenging task due to the diversity of tissue structures in the histopathological images. In this work, a novel WSI-based multi-dictionaries learning framework was proposed to adaptively extract the effective features of WSI for prognostic analysis of ccRCC. This framework included multi-dictionaries learning stage based on patch level and survival model construction stage based on patient level. The proposed model was evaluated on 378 hematoxylin-eosin stained WSIs form Cancer Genome Atlas database (TCGA-KIRC). The C-index was 0.681, and AUC was 0.751(P＜0.05). Compared with the traditional Boosted model and Random Survival Trees model, the improvements on C-index were respectively 0.138 and 0.155, and the improvements on AUC was respectively 0.149 and 0.191. Compared with the two deep learning model (DeepSurv and WSISA), the improvements on C-index were respectively 0.046 and 0.035, and the improvements on AUC was respectively 0.096 and 0.090. The results showed that the proposed model achieved superior performance for prognostic analysis of renal clear cell carcinoma.
涂超, 宁振源, 张煜. 基于多字典学习框架的肾透明细胞癌预后分析模型[J]. 中国生物医学工程学报, 2021, 40(4): 385-393.
Tu Chao, Ning Zhenyuan, Zhang Yu. Prognostic Analysis Model of Renal Clear Cell Carcinoma Based on Multi-Dictionary Learning. Chinese Journal of Biomedical Engineering, 2021, 40(4): 385-393.
[1] Zhou Ming, He Huiying. Renal Cell Carcinoma: Clinical Management [M]. Campbell and Rini: Humana Press, 2013: 23-41. [2] Hsieh JJ, Purdue MP, Signoretti S, et al. Renal cell carcinoma [J]. Nat Rev Dis Primers, 2017, 3: 17009. [3] Network Cancer Genome Atlas Research. Comprehensive molecular characterization of clear cell renal cell carcinoma [J]. Nature, 2013, 499(7456): 43-49. [4] Siegel RL, Miller KD, Jemal Ahmedin. Cancer statistics, 2019 [J]. CA: A Cancer Journal for Clinicians, 2019, 69(1): 7-34. [5] Gulati S, Martinez P, Joshi T, et al. Systematic evaluation of the prognostic impact and intratumour heterogeneity of clear cell renal cell carcinoma biomarkers [J]. European Urology, 2014, 66(5): 936-948. [6] Kim SP, Alt AL, Weight CJ, et al. Independent validation of the 2010 American Joint Committee on Cancer TNM classification for renal cell carcinoma: Results from a large, single institution cohort [J]. The Journal of Urology, 2011, 185(6): 2035-2039. [7] Zisman A, Pantuck AJ, Dorey F, et al. Improved prognostication of renal cell carcinoma using an integrated staging system [J]. Journal of Clinical Oncology, 2001, 19(6): 1649-1657. [8] Frank I, Blute ML, Cheville JC, et al. An outcome prediction model for patients with clear cell renal cell carcinoma treated with radical nephrectomy based on tumor stage, size, grade and necrosis: The SSIGN score [J]. The Journal of urology, 2002, 168(6): 2395-2400. [9] Leibovich BC, Blute ML, Cheville JC, et al. Prediction of progression after radical nephrectomy for patients with clear cell renal cell carcinoma: A stratification tool for prospective clinical trials [J]. Cancer: Interdisciplinary International Journal of the American Cancer Society, 2003, 97(7): 1663-1671. [10] Ning Zhenyuan, Pan Weihao, Chen Yuting, et al. Integrative analysis of cross-modal features for the prognosis prediction of clear cell renal cell carcinoma [J]. Bioinformatics, 2020, 36(9): 2888-2895. [11] Cheng Jun, Mo Xiaokui, Wang Xusheng, et al. Identification of topological features in renal tumor microenvironment associated with patient survival [J]. Bioinformatics, 2018, 34(6): 1024-1030. [12] Belsare AD, Mushrif MM. Histopathological image analysis using image processing techniques: An overview [J]. Signal & Image Processing: An International Journal, 2012, 3(4): 23-36. [13] Li Zhongyu, Zhang Xiaofan, Müller Henning, et al. Large-scale retrieval for medical image analytics: A comprehensive review [J]. Medical Image Analysis, 2018, 43: 66-84. [14] Lopez CM, Agaian S. A new set of wavelet-and fractals-based features for Gleason grading of prostate cancer histopathology images [C]// Image Processing: Algorithms and Systems XI. San Diego: International Society for Optics and Photonics, 2013, 8655: 865516. [15] Doyle S, Agner S, Madabhushi A, et al. Automated grading of breast cancer histopathology using spectral clustering with textural and architectural image features [C]//The 5th IEEE International Symposium on Biomedical Imaging. Piscataway: IEEE, 2008: 496-499. [16] Ozdemir E, Gunduz-Demir C. A hybrid classification model for digital pathology using structural and statistical pattern recognition [J]. IEEE Transactions on Medical Imaging, 2012, 32(2): 474-483. [17] Caicedo JC, Cruz A, Gonzalez FA. Histopathology image classification using bag of features and kernel functions [C]// Conference on Artificial Intelligence in Medicine in Europe. Berlin: Springer, 2009: 126-135. [18] Li Xuelu, Monga V, Rao UKA. Analysis-synthesis learning with shared features: Algorithms for histology image classification [J]. IEEE Transactions on Biomedical Engineering, 2020, 67(4): 1061-1073. [19] LeCun Y, Bengio Y, Hinton G. Deep learning [J]. Nature, 2015, 521(7553): 436-444. [20] Litjens G, Kooi T, Bejnordi BE, et al. A survey on deep learning in medical image analysis [J]. Med Image Anal, 2017, 42: 60-88. [21] Liu Yu, Chen Xun, Wang Zengfu, et al. Deep learning for pixel-level image fusion: Recent advances and future prospects [J]. Information Fusion, 2018, 42: 158-173. [22] Vu TH, Mousavi HS, Monga V, et al. Histopathological image classification using discriminative feature-oriented dictionary learning [J]. IEEE Transactions on Medical Imaging, 2016, 35(3): 738-751. [23] Wu Guoqing, Chen Yinsheng, Wang Yuanyuan, et al. Sparse representation-based radiomics for the diagnosis of brain tumors [J]. IEEE Trans Med Imaging, 2018, 37(4): 893-905. [24] Otsu N. A threshold selection method from gray-level histograms [J]. IEEE Transactions on Systems, Man, and Cybernetics, 1979, 9(1): 62-66. [25] Aharon M, Elad M, Bruckstein A. K-SVD: An algorithm for designing overcomplete dictionaries for sparse representation [J]. IEEE Transactions on Signal Processing, 2006, 54(11): 4311-4322. [26] Cox DR. Regression models and life‐tables [J]. Journal of the Royal Statistical Society: Series B (Methodological), 1972, 34(2): 187-202. [27] Friedman JH. Stochastic gradient boosting [J]. Computational Statistics & Data Analysis, 2002, 38(4): 367-378. [28] Ishwaran H, Kogalur UB, Blackstone EH, et al. Random survival forests [J]. The Annals of Applied Statistics, 2008, 2(3): 841-860. [29] Katzman JL, Shaham U, Cloninger A, et al. DeepSurv: Personalized treatment recommender system using a Cox proportional hazards deep neural network [J]. BMC Medical Research Methodology, 2018, 18(1): 24. [30] Zhu Xinliang, Yao Jiawen., Zhu Feiyun., et al. WSISA: Making survival prediction from whole slide histopathological images [C]// IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Piscataway: IEEE, 2017: 6855-6863. [31] Cheng Jun, Mo Xiaokui, Wang Xusheng, et al. Identification of topological features in renal tumor microenvironment associated with patient survival [J]. Bioinformatics, 2017, 34(6): 1024-1030. [32] 张丹, 陆铖, 代才, 等. 基于数字组织病理图像的肿瘤与微环境相互作用风险模型在口咽癌症中的预后研究 [J]. 中国生物医学工程学报, 2020, 39(5): 541-549.
