Share：
Share this content in WeChat
X
[Chinese][PDF]4275
• Clinical Article •
Preoperative prediction of pathological grading in bladder cancer by deep residual network model based on T2WI
HUANG Xiang  CAO Kangyang  ZOU Yujian  DENG Lei  ZHANG Weijing  YANG Shuiqing  ZHANG Kunlin  ZHU Yurong  LI Jianpeng 

Cite this article as: HUANG X, CAO K Y, ZOU Y J, et al. Preoperative prediction of pathological grading in bladder cancer by deep residual network model based on T2WI[J]. Chin J Magn Reson Imaging, 2024, 15(1): 125-131. DOI:10.12015/issn.1674-8034.2024.01.020.


[Abstract] Objective To develop and validate the efficacy of a deep learning (DL) model by 50-layer deep residual network (ResNet-50) based on T2WI for preoperative prediction of preoperative pathological grading of bladder cancer (BCa).Materials and Methods A total of 211 tumors in 169 BCa patients [109 for training and 47 for internal test, from the Tenth Affiliated Hospital of Southern Medical University (centre 1); 55 for external test, from Sun Yat-sen University Cancer centre (centre 2)] were enrolled, including 111 tumors of high grade uroepithelial carcinoma (HGUC) and 100 tumors of low grade uroepithelial carcinoma (LGUC). Grade determination was confirmed by pathological examination. ResNet-50 was used to construct the models based on the internal training set from centre 1. The optimal model was selected from the resulting models after being tested on the internal test set from centre 1 and validated independently on the external test set from centre 2. The performance of the model was evaluated using the area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, and specificity, with feature visualization images presented.Results In the internal test set, we achieved an AUC of 0.856 [95% confidence interval (CI): 0.723-0.941], accuracy of 80.9% (95% CI: 69.6%-92.1%), sensitivity of 77.8% (95% CI: 65.9%-89.7%), and specificity of 82.8% (95% CI: 72.0%-93.6%). In the external test set, we achieved an AUC of 0.814 (95% CI: 0.686-0.906), accuracy of 78.2% (95% CI: 67.3%-89.1%), sensitivity of 77.3% (95% CI: 66.2%-88.3%), and specificity of 81.8% (95% CI: 71.6%-92.0%). Feature visualization showed that the activated regions overlapped with the BCa lesions largely, indicating the DL model identified the target area of BCa correctly. And the t-distributed stochastic neighbor embedding (T-SNE) helped to distinguish HGUC from LGUC in a certain extent.Conclusions This study is the first to establish a preoperative BCa pathological grading prediction model based on T2WI using DL methods and be validated across two centres. With high prediction accuracy, the model is non-invasive, objective, and has strong repeatability and generalization performance, which contribute to more accurate clinical preoperative diagnosis.
[Keywords] bladder neoplasms；deep learning；magnetic resonance imaging；neoplasm grading

HUANG Xiang1   CAO Kangyang2   ZOU Yujian1   DENG Lei1   ZHANG Weijing3   YANG Shuiqing1   ZHANG Kunlin1   ZHU Yurong4   LI Jianpeng1*  

1 Department of Radiology, the Tenth Affiliated Hospital of Southern Medical University (Dongguan People's hospital), Dongguan 523000, China

2 Medical AI Lab, School of Biomedical Engineering, Health Science Centre, Shenzhen University, Shenzhen 518000, China

3 Department of Radiology, Sun Yat-sen University Cancer Centre, Guangzhou 510000, China

4 Department of Pathology, the Tenth Affiliated Hospital of Southern Medical University (Dongguan People's hospital), Dongguan 523000, China

Corresponding author: LI J P, E-mail: ljp0885@qq.com

Conflicts of interest   None.

Received  2023-04-15
Accepted  2023-11-03
DOI: 10.12015/issn.1674-8034.2024.01.020
Cite this article as: HUANG X, CAO K Y, ZOU Y J, et al. Preoperative prediction of pathological grading in bladder cancer by deep residual network model based on T2WI[J]. Chin J Magn Reson Imaging, 2024, 15(1): 125-131. DOI:10.12015/issn.1674-8034.2024.01.020.

[1]
SUNG H, FERLAY J, SIEGEL R L, et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries[J]. CA Cancer J Clin, 2021, 71(3): 209-249. DOI: 10.3322/caac.21660.
[2]
CAO W, CHEN H D, YU Y W, et al. Changing profiles of cancer burden worldwide and in China: a secondary analysis of the global cancer statistics 2020[J]. Chin Med J, 2021, 134(7): 783-791. DOI: 10.1097/CM9.0000000000001474.
[3]
HUMPHREY P A, MOCH H, CUBILLA A L, et al. The 2016 WHO classification of tumours of the urinary system and male genital organs-part B: prostate and bladder tumours[J]. Eur Urol, 2016, 70(1): 106-119. DOI: 10.1016/j.eururo.2016.02.028.
[4]
POWLES T, BELLMUNT J, COMPERAT E, et al. Bladder cancer: ESMO Clinical Practice Guideline for diagnosis, treatment and follow-up[J]. Ann Oncol, 2022, 33(3): 244-258. DOI: 10.1016/j.annonc.2021.11.012.
[5]
BABJUK M, BURGER M, COMPÉRAT E M, et al. European association of urology guidelines on non-muscle-invasive bladder cancer (TaT1 and carcinoma in situ) - 2019 update[J]. Eur Urol, 2019, 76(5): 639-657. DOI: 10.1016/j.eururo.2019.08.016.
[6]
PANEBIANCO V, NARUMI Y, ALTUN E, et al. Multiparametric magnetic resonance imaging for bladder cancer: development of VI-RADS (vesical imaging-reporting and data system)[J]. Eur Urol, 2018, 74(3): 294-306. DOI: 10.1016/j.eururo.2018.04.029.
[7]
TURKER P, BOSTROM P J, WROCLAWSKI M L, et al. Upstaging of urothelial cancer at the time of radical cystectomy: factors associated with upstaging and its effect on outcome[J]. BJU Int, 2012, 110(6): 804-811. DOI: 10.1111/j.1464-410X.2012.10939.x.
[8]
SHARIAT S F, PALAPATTU G S, KARAKIEWICZ P I, et al. Discrepancy between clinical and pathologic stage: impact on prognosis after radical cystectomy[J]. Eur Urol, 2007, 51(1): 137-151. DOI: 10.1016/j.eururo.2006.05.021.
[9]
Guidelines Working Committee of Chinese Society of Clinical Oncology. Guidelines of Chinese society of clinical oncology (CSCO) urothelial cancer[M]. Beijing: People's Medical Publishing House, 2022: 440-441.
[10]
FLAIG T W, SPIESS P E, ABERN M, et al. NCCN guidelines® insights: bladder cancer, version 2.2022[J]. J Natl Compr Canc Netw, 2022, 20(8): 866-878. DOI: 10.6004/jnccn.2022.0041.
[11]
DENG L, ZOU Y J, YANG S Q, et al. VI-RADS based on multi-parametric MRI for the prediction of muscle-invasion in bladder cancer and MRI findings of muscle-invasion in bladder cancer located in ureteral orifice[J]. Chin J Magn Reson Imag, 2022, 13(7): 121-125. DOI: 10.12015/issn.1674-8034.2022.07.022.
[12]
WANG H J, HU D K, YAO H H, et al. Radiomics analysis of multiparametric MRI for the preoperative evaluation of pathological grade in bladder cancer tumors[J]. Eur Radiol, 2019, 29(11): 6182-6190. DOI: 10.1007/s00330-019-06222-8.
[13]
ZHANG X X, WANG Y C, WANG S C, et al. Feasibility study of deep learning reconstruction in the clinical application of MRI in bladder cancer[J]. Chin J Magn Reson Imag, 2023, 14(5): 36-40. DOI: 10.12015/issn.1674-8034.2023.05.008.
[14]
ZOU Y, CAI L K, CHEN C X, et al. Multi-task deep learning based on T2-Weighted Images for predicting Muscular-Invasive Bladder Cancer[J/OL]. Comput Biol Med, 2022, 151(Pt A): 106219 [2023-02-12]. https://doi.org/10.1016/j.compbiomed.2022.106219. DOI: 10.1016/j.compbiomed.2022.106219.
[15]
LI J P, CAO K Y, LIN H X, et al. Predicting muscle invasion in bladder cancer by deep learning analysis of MRI: comparison with vesical imaging-reporting and data system[J]. Eur Radiol, 2023, 33(4): 2699-2709. DOI: 10.1007/s00330-022-09272-7.
[16]
LI J P, QIU Z X, CAO K Y, et al. Predicting muscle invasion in bladder cancer based on MRI: A comparison of radiomics, and single-task and multi-task deep learning[J/OL]. Comput Methods Programs Biomed, 2023, 233: 107466 [2023-03-22]. https://doi.org/10.1016/j.cmpb.2023.107466. DOI: 10.1016/j.cmpb.2023.107466.
[17]
LEO M C, MCMULLEN C K, O'KEEFFE-ROSETTI M, et al. External validation of the EORTC and NCCN bladder cancer recurrence and progression risk calculators in a U.S. community-based health system[J/OL]. Urol Oncol, 2020, 38(2): 39.e21-39.e27 [2022-11-16]. https://doi.org/10.1016/j.urolonc.2019.10.003. DOI: 10.1016/j.urolonc.2019.10.003.
[18]
WANG H J, PUI M H, GUO Y, et al. Multiparametric 3-T MRI for differentiating low-versus high-grade and category T1 versus T2 bladder urothelial carcinoma[J]. AJR Am J Roentgenol, 2015, 204(2): 330-334. DOI: 10.2214/AJR.14.13147.
[19]
SVATEK R S, SHARIAT S F, NOVARA G, et al. Discrepancy between clinical and pathological stage: external validation of the impact on prognosis in an international radical cystectomy cohort[J]. BJU Int, 2011, 107(6): 898-904. DOI: 10.1111/j.1464-410X.2010.09628.x.
[20]
MARIAPPAN P, ZACHOU A, GRIGOR K M, et al. Detrusor muscle in the first, apparently complete transurethral resection of bladder tumour specimen is a surrogate marker of resection quality, predicts risk of early recurrence, and is dependent on operator experience[J]. Eur Urol, 2010, 57(5): 843-849. DOI: 10.1016/j.eururo.2009.05.047.
[21]
TAKEUCHI M, SASAKI S, ITO M, et al. Urinary bladder cancer: diffusion-weighted MR imaging: accuracy for diagnosing T stage and estimating histologic grade[J]. Radiology, 2009, 251(1): 112-121. DOI: 10.1148/radiol.2511080873.
[22]
GUPTA N, SUREKA B, KUMAR M M, et al. Comparison of dynamic contrast-enhanced and diffusion weighted magnetic resonance image in staging and grading of carcinoma bladder with histopathological correlation[J]. Urol Ann, 2015, 7(2): 199-204. DOI: 10.4103/0974-7796.150480.
[23]
WANG Y C, SHEN Y Q, HU X M, et al. Application of R2* and apparent diffusion coefficient in estimating tumor grade and T category of bladder cancer[J]. AJR Am J Roentgenol, 2020, 214(2): 383-389. DOI: 10.2214/AJR.19.21668.
[24]
CLUCERU J, INTERIAN Y, PHILLIPS J J, et al. Improving the noninvasive classification of glioma genetic subtype with deep learning and diffusion-weighted imaging[J]. Neuro Oncol, 2022, 24(4): 639-652. DOI: 10.1093/neuonc/noab238.
[25]
XU Y, ZHOU C Y, HE X J, et al. Deep learning-assisted LI-RADS grading and distinguishing hepatocellular carcinoma (HCC) from non-HCC based on multiphase CT: a two-center study[J/OL]. Eur Radiol, 2023 [2023-07-03]. https://link.springer.com/article/10.1007/s00330-023-09857-w. DOI: 10.1007/s00330-023-09857-w.
[26]
VAN DER VOORT S R, INCEKARA F, WIJNENGA M M J, et al. Combined molecular subtyping, grading, and segmentation of glioma using multi-task deep learning[J]. Neuro Oncol, 2023, 25(2): 279-289. DOI: 10.1093/neuonc/noac166.
[27]
ZHANG T H, LONG X, WANG L Q, et al. The value of prediction model based on MRI radiomics in evaluating the histological grade of urothelial carcinoma fo bladder[J]. J Clin Radiol, 2022, 41(1): 116-120. DOI: 10.13437/j.cnki.jcr.2022.01.022.
[28]
SENG H S, ARGHA A, LIAW S T, et al. Machine and deep learning for tuberculosis detection on chest X-rays: systematic literature review[J/OL]. J Med Internet Res, 2023, 25: e43154 [2023-07-03]. https://www.jmir.org/2023/1/e43154. DOI: 10.2196/43154.
[29]
BRESSEM K K, ADAMS L C, ERXLEBEN C, et al. Comparing different deep learning architectures for classification of chest radiographs[J/OL]. Sci Rep, 2020, 10(1): 13590 [2023-07-04]. https://www.nature.com/articles/s41598-020-70479-z. DOI: 10.1038/s41598-020-70479-z.
[30]
TALO M, YILDIRIM O, BALOGLU U B, et al. Convolutional neural networks for multi-class brain disease detection using MRI images[J/OL]. Comput Med Imaging Graph, 2019, 78: 101673 [2023-07-03]. https://doi.org/10.1016/j.compmedimag.2019. DOI: 10.1016/j.compmedimag.2019.101673.

PREV Value of ZOOM-mDixon-derived T2*/R2* imaging in preoperative predicting lymph node metastasis in pancreatic ductal adenocarcinoma
NEXT The value of APT imaging in evaluating the therapeutic efficacy of concurrent chemoradiotherapy for cervical cancer
  


LINK


Tel & Fax: +8610-67113815    E-mail: editor@cjmri.cn