Skip to main content

Advertisement

SpringerLink
Log in

An interpretable machine learning model based on contrast-enhanced CT parameters for predicting treatment response to conventional transarterial chemoembolization in patients with hepatocellular carcinoma

  • Abdominal Radiology
  • Published:
La radiologia medica Aims and scope Submit manuscript

Abstract

Objective

To explore the potential of pre-therapy computed tomography (CT) parameters in predicting the treatment response to initial conventional TACE (cTACE) in intermediate-stage hepatocellular carcinoma (HCC) and develop an interpretable machine learning model.

Methods

This retrospective study included 367 patients with intermediate-stage HCC who received cTACE as first-line therapy from three centers. We measured the mean attenuation values of target lesions on multi-phase contrast-enhanced CT and further calculated three CT parameters, including arterial (AER), portal venous (PER), and arterial portal venous (APR) enhancement ratios. We used logistic regression analysis to select discriminative features and trained three machine learning models via 5-fold cross-validation. The performance in predicting treatment response was evaluated in terms of discrimination, calibration, and clinical utility. Afterward, a Shapley additive explanation (SHAP) algorithm was leveraged to interpret the outputs of the best-performing model.

Results

The mean diameter, ECOG performance status, and cirrhosis were the important clinical predictors of cTACE treatment response, by multiple logistic regression. Adding the CT parameters to clinical variables showed significant improvement in performance (net reclassification index, 0.318, P < 0.001). The Random Forest model (hereafter, RF-combined model) integrating CT parameters and clinical variables demonstrated the highest performance on external validation dataset (AUC of 0.800). The decision curve analysis illustrated the optimal clinical benefits of RF-combined model. This model could successfully stratify patients into responders and non-responders with distinct survival (P = 0.001).

Conclusion

The RF-combined model can serve as a robust and interpretable tool to identify the appropriate crowd for cTACE sessions, sparing patients from receiving ineffective and unnecessary treatments.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5

Similar content being viewed by others

Abbreviations

AER:

Arterial enhancement ratio

PER:

Portal venous enhancement ratio

APR:

Arterial portal venous enhancement ratio

CT:

Computed tomography

cTACE:

Conventional transarterial chemoembolization

ML:

Machine learning

HCC:

Hepatocellular carcinoma

mRECIST:

Modified Response Evaluation Criteria in Solid Tumors

LR:

Logistic regression

RF:

Random forest

SVM:

Support vector machine

SHAP:

Shapley additive explanation

References

  1. Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, Bray F (2021) Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 71(3):209–249. https://doi.org/10.3322/caac.21660

    Article  PubMed  Google Scholar 

  2. Heimbach JK, Kulik LM, Finn RS, Sirlin CB, Abecassis MM, Roberts LR, Zhu AX, Murad MH, Marrero JA (2018) AASLD guidelines for the treatment of hepatocellular carcinoma. Hepatology 67(1):358–380. https://doi.org/10.1002/hep.29086

    Article  PubMed  Google Scholar 

  3. EASL (2018) EASL clinical practice guidelines: management of hepatocellular carcinoma. J Hepatol 69(1):182–236. https://doi.org/10.1016/j.jhep.2018.03.019

    Article  Google Scholar 

  4. Bolondi L, Burroughs A, Dufour JF, Galle PR, Mazzaferro V, Piscaglia F, Raoul JL, Sangro B (2012) Heterogeneity of patients with intermediate (BCLC B) hepatocellular carcinoma: proposal for a subclassification to facilitate treatment decisions. Semin Liver Dis 32(4):348–359. https://doi.org/10.1055/s-0032-1329906

    Article  CAS  PubMed  Google Scholar 

  5. Han K, Kim JH (2015) Transarterial chemoembolization in hepatocellular carcinoma treatment: Barcelona clinic liver cancer staging system. World J Gastroenterol 21(36):10327–10335. https://doi.org/10.3748/wjg.v21.i36.10327

    Article  CAS  PubMed  PubMed Central  ADS  Google Scholar 

  6. Ma W, Jia J, Wang S, Bai W, Yi J, Bai M, Quan Z, Yin Z, Fan D, Wang J, Han G (2014) The prognostic value of 18F-FDG PET/CT for hepatocellular carcinoma treated with transarterial chemoembolization (TACE). Theranostics 4(7):736–744. https://doi.org/10.7150/thno.8725

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. Pinato DJ, Sharma R, Allara E, Yen C, Arizumi T, Kubota K, Bettinger D, Jang JW, Smirne C, Kim YW, Kudo M, Howell J, Ramaswami R, Burlone ME, Guerra V, Thimme R, Ishizuka M, Stebbing J, Pirisi M, Carr BI (2017) The ALBI grade provides objective hepatic reserve estimation across each BCLC stage of hepatocellular carcinoma. J Hepatol 66(2):338–346. https://doi.org/10.1016/j.jhep.2016.09.008

    Article  PubMed  Google Scholar 

  8. Sieghart W, Hucke F, Pinter M, Graziadei I, Vogel W, Müller C, Heinzl H, Trauner M, Peck-Radosavljevic M (2013) The ART of decision making: retreatment with transarterial chemoembolization in patients with hepatocellular carcinoma. Hepatology 57(6):2261–2273. https://doi.org/10.1002/hep.26256

    Article  CAS  PubMed  Google Scholar 

  9. Adhoute X, Penaranda G, Naude S, Raoul JL, Perrier H, Bayle O, Monnet O, Beaurain P, Bazin C, Pol B, Folgoc GL, Castellani P, Bronowicki JP, Bourlière M (2015) Retreatment with TACE: the ABCR SCORE, an aid to the decision-making process. J Hepatol 62(4):855–862. https://doi.org/10.1016/j.jhep.2014.11.014

    Article  PubMed  Google Scholar 

  10. Boas FE, Brody LA, Erinjeri JP, Yarmohammadi H, Shady W, Kishore S, Sofocleous CT (2016) Quantitative measurements of enhancement on preprocedure triphasic CT Can predict response of colorectal liver metastases to radioembolization. AJR Am J Roentgenol 207(3):671–675. https://doi.org/10.2214/AJR.15.15767

    Article  PubMed  PubMed Central  Google Scholar 

  11. Liu Y, Chen W, Cui W, Liu H, Zhou X, Chen L, Li J, Chen M, Chen J, Wang Y (2020) Quantitative pretreatment CT parameters as predictors of tumor response of neuroendocrine tumor liver metastasis to transcatheter arterial bland embolization. Neuroendocrinology 110(7–8):697–704. https://doi.org/10.1159/000504257

    Article  CAS  PubMed  Google Scholar 

  12. Marrache F, Vullierme MP, Roy C, El Assoued Y, Couvelard A, O’Toole D, Mitry E, Hentic O, Hammel P, Lévy P, Ravaud P, Rougier P, Ruszniewski P (2007) Arterial phase enhancement and body mass index are predictors of response to chemoembolisation for liver metastases of endocrine tumours. Br J Cancer 96(1):49–55. https://doi.org/10.1038/sj.bjc.6603526

    Article  CAS  PubMed  Google Scholar 

  13. Syha R, Grözinger G, Grosse U, Maurer M, Zender L, Horger M, Nikolaou K, Ketelsen D (2016) Parenchymal blood volume assessed by C-arm-based computed tomography in immediate posttreatment evaluation of drug-eluting bead transarterial chemoembolization in hepatocellular carcinoma. Invest Radiol 51(2):121–126. https://doi.org/10.1097/rli.0000000000000215

    Article  CAS  PubMed  Google Scholar 

  14. Liu D, Liu F, Xie X, Su L, Liu M, Xie X, Kuang M, Huang G, Wang Y, Zhou H, Wang K, Lin M, Tian J (2020) Accurate prediction of responses to transarterial chemoembolization for patients with hepatocellular carcinoma by using artificial intelligence in contrast-enhanced ultrasound. Eur Radiol 30(4):2365–2376. https://doi.org/10.1007/s00330-019-06553-6

    Article  PubMed  Google Scholar 

  15. Kong C, Zhao Z, Chen W, Lv X, Shu G, Ye M, Song J, Ying X, Weng Q, Weng W, Fang S, Chen M, Tu J, Ji J (2021) Prediction of tumor response via a pretreatment MRI radiomics-based nomogram in HCC treated with TACE. Eur Radiol 31(10):7500–7511. https://doi.org/10.1007/s00330-021-07910-0

    Article  PubMed  PubMed Central  Google Scholar 

  16. Peng J, Kang S, Ning Z, Deng H, Shen J, Xu Y, Zhang J, Zhao W, Li X, Gong W, Huang J, Liu L (2020) Residual convolutional neural network for predicting response of transarterial chemoembolization in hepatocellular carcinoma from CT imaging. Eur Radiol 30(1):413–424. https://doi.org/10.1007/s00330-019-06318-1

    Article  PubMed  Google Scholar 

  17. Lundberg SM, Nair B, Vavilala MS, Horibe M, Eisses MJ, Adams T, Liston DE, Low DK, Newman SF, Kim J, Lee SI (2018) Explainable machine-learning predictions for the prevention of hypoxaemia during surgery. Nat Biomed Eng 2(10):749–760. https://doi.org/10.1038/s41551-018-0304-0

    Article  PubMed  PubMed Central  Google Scholar 

  18. Caruana R, Lou Y, Gehrke J, Koch P, Sturm M, Elhadad N (2015) Intelligible models for healthcare: predicting pneumonia risk and hospital 30-day readmission. Paper presented at the proceedings of the 21th ACM SIGKDD international conference on knowledge discovery and data mining, Sydney, NSW, Australia

  19. Lencioni R, Llovet JM (2010) Modified RECIST (mRECIST) assessment for hepatocellular carcinoma. Semin Liver Dis 30(1):52–60. https://doi.org/10.1055/s-0030-1247132

    Article  CAS  PubMed  Google Scholar 

  20. Weir JP (2005) Quantifying test-retest reliability using the intraclass correlation coefficient and the SEM. J Strength Cond Res 19(1):231–240. https://doi.org/10.1519/15184.1

    Article  PubMed  Google Scholar 

  21. Liu Y, Chen W, Cui W, Liu H, Zhou X, Chen L, Li J, Chen M, Chen J, Wang Y (2019) Quantitative pretreatment CT parameters as predictors of tumor response of NET liver metastasis to TAE. Neuroendocrinology. https://doi.org/10.1159/000504257

    Article  PubMed  Google Scholar 

  22. Lundberg SM, Erion G, Chen H, DeGrave A, Prutkin JM, Nair B, Katz R, Himmelfarb J, Bansal N, Lee S-I (2020) From local explanations to global understanding with explainable AI for trees. Nat Mach Intell 2(1):56–67. https://doi.org/10.1038/s42256-019-0138-9

    Article  PubMed  PubMed Central  Google Scholar 

  23. Zhang L, Jiang Y, Jin Z, Jiang W, Zhang B, Wang C, Wu L, Chen L, Chen Q, Liu S, You J, Mo X, Liu J, Xiong Z, Huang T, Yang L, Wan X, Wen G, Han XG, Fan W, Zhang S (2022) Real-time automatic prediction of treatment response to transcatheter arterial chemoembolization in patients with hepatocellular carcinoma using deep learning based on digital subtraction angiography videos. Cancer Imaging. https://doi.org/10.1186/s40644-022-00457-3

    Article  PubMed  PubMed Central  Google Scholar 

  24. Berenguer R, Pastor-Juan MDR, Canales-Vázquez J, Castro-García M, Villas MV, Mansilla Legorburo F, Sabater S (2018) Radiomics of CT features may be nonreproducible and redundant: influence of CT acquisition parameters. Radiology 288(2):407–415. https://doi.org/10.1148/radiol.2018172361

    Article  PubMed  Google Scholar 

  25. Selvaraju RR, Cogswell M, Das A, Vedantam R, Parikh D, Batra D Grad-cam: visual explanations from deep networks via gradient-based localization. In: Proceedings of the IEEE international conference on computer vision, 2017. pp 618–626

  26. Fronda M, Doriguzzi Breatta A, Gatti M, Calandri M, Maglia C, Bergamasco L, Righi D, Faletti R, Fonio P (2021) Quantitative assessment of HCC wash-out on CT is a predictor of early complete response to TACE. Eur Radiol 31(9):6578–6588. https://doi.org/10.1007/s00330-021-07792-2

    Article  PubMed  PubMed Central  Google Scholar 

  27. Rokach L, Maimon O (2005) Decision trees. IEEE Trans Syst Man Cybern Part C 35(4):476–487. https://doi.org/10.1007/0-387-25465-X_9

    Article  Google Scholar 

  28. Golden CE, Rothrock MJ Jr, Mishra A (2019) Comparison between random forest and gradient boosting machine methods for predicting Listeria spp. prevalence in the environment of pastured poultry farms. Food Res Int 122:47–55. https://doi.org/10.1016/j.foodres.2019.03.062

    Article  PubMed  Google Scholar 

  29. Doupe P, Faghmous J, Basu S (2019) Machine learning for health services researchers. Value Health 22(7):808–815. https://doi.org/10.1016/j.jval.2019.02.012

    Article  PubMed  Google Scholar 

  30. Li K, Yao S, Zhang Z, Cao B, Wilson CM, Kalos D, Kuan PF, Zhu R, Wang X (2022) Efficient gradient boosting for prognostic biomarker discovery. Bioinformatics. https://doi.org/10.1093/bioinformatics/btab869

    Article  PubMed  PubMed Central  Google Scholar 

  31. Thorsen-Meyer H-C, Nielsen AB, Nielsen AP, Kaas-Hansen BS, Toft P, Schierbeck J, Strøm T, Chmura PJ, Heimann M, Dybdahl L, Spangsege L, Hulsen P, Belling K, Brunak S, Perner A (2020) Dynamic and explainable machine learning prediction of mortality in patients in the intensive care unit: a retrospective study of high-frequency data in electronic patient records. Lancet Digit Health 2(4):e179–e191. https://doi.org/10.1016/S2589-7500(20)30018-2

    Article  PubMed  Google Scholar 

  32. Lundberg SM, Nair B, Vavilala MS, Horibe M, Eisses MJ, Adams T, Liston DE, Low DK-W, Newman S-F, Kim J, Lee S-I (2018) Explainable machine-learning predictions for the prevention of hypoxaemia during surgery. Nat Biomed Eng 2(10):749–760. https://doi.org/10.1038/s41551-018-0304-0

    Article  PubMed  PubMed Central  Google Scholar 

  33. Lu J, Zhang S, Jiang H, Yang L, Hao D, Yang Y, Li X, Chen A, Shao J, Liu X (2021) Gestational diabetes mellitus prediction model: A risk factor analysis of pregnant women with gestational diabetes mellitus and a normal oral glucose tolerance test in the second trimester of pregnancy. Technol Health Care Off J Eur Soc Eng Med. https://doi.org/10.3233/THC-218006

    Article  Google Scholar 

  34. Zhang S, Zhao B, Chen D, Qi Y, Ma Y, Ma J, Xie W, Guo H (2021) Anesthetic management of precise radiotherapy under apnea-like condition. J Int Med Res 49(3):300060521990260. https://doi.org/10.1177/0300060521990260

    Article  PubMed  Google Scholar 

Download references

Funding

This research was supported by grants from the National Key Research and Development Program of China (2023YFF1204600); the National Natural Science Foundation of China (82227802, 82302306); the Clinical Frontier Technology Program of the First Affiliated Hospital of Jinan University, China (No. JNU1AF-CFTP-2022-a01201); the Science and Technology Projects in Guangzhou (202201020022, 2023A03J1036, 2023A03J1038); the Science and Technology Youth Talent Nurturing Program of Jinan University (21623209); and the Postdoctoral Research Foundation of China (2022M721349).

Author information

Author notes

  1. Lu Zhang, Zhe Jin, Chen Li, and Zicong He have contributed equally to this work.

Authors and Affiliations

  1. Department of Radiology, The First Affiliated Hospital of Jinan University, No. 613 Huangpu West Road, Tianhe District, Guangzhou, 510627, Guangdong, China

    Lu Zhang, Zhe Jin, Chen Li, Zicong He, Bin Zhang, Qiuying Chen, Jingjing You, Xiao Ma, Hui Shen, Fei Wang & Shuixing Zhang

  2. Department of Interventional Therapy, The Second Affiliated Hospital of Guangzhou University of Traditional Chinese Medicine Guangzhou, Guangdong, 510627, China

    Lingeng Wu

  3. Department of Radiology, The People’s Hospital of Wenshan Prefecture, No. 228 Kaihua East Road, Wenshan, 663000, Yunnan, China

    Cunwen Ma

Authors
  1. Lu Zhang
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Zhe Jin
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Chen Li
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Zicong He
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Bin Zhang
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Qiuying Chen
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Jingjing You
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Xiao Ma
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Hui Shen
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. Fei Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

  11. Lingeng Wu
    View author publications

    You can also search for this author in PubMed Google Scholar

  12. Cunwen Ma
    View author publications

    You can also search for this author in PubMed Google Scholar

  13. Shuixing Zhang
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

LZ, LW, CM, ZH, and SZ conceived and designed the project; ZJ, CL, BZ, QC, ZH, JY, XM, and HS performed the research and collected the data; LW and CL analyzed the data; and LZ, ZJ, ZH, FW, and CM wrote the paper. All authors read and approved the final manuscript.

Corresponding authors

Correspondence to Lingeng Wu, Cunwen Ma or Shuixing Zhang.

Ethics declarations

Competing interests

The authors have no relevant financial or non-financial interests to disclose.

Ethical approval

This retrospective study was performed in line with the principles of the Declaration of Helsinki. Approval was granted by the Ethics Committee of the institutional Ethics Committees of the First Affiliated Hospital of Jinan University, and the requirement for informed consent was waived.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary Figure S1

. Performance of different models on the cross-validation cohort. The following order: A was LR-clinical model, B was LR-combined model, C was SVM-clinical model, D was SVM-combined model, E was RF-clinical model, F was RF-combined model. The performance of RF-combined model was the best among models in the cross-validation cohort. LR, Logistic regression; SVM, support vector machine; RF, random forest. (TIF 3099 kb)

Supplementary Figure S2.

The learning curve of the different models in the training and validation cohorts. The following order: A was LR-clinical model, B was LR-combined model, C was SVM-clinical model, D was SVM-combined model, E was RF-clinical model, F was RF-combined model. LR, Logistic regression; SVM, support vector machine; RF, random forest. (TIF 226 kb)

Supplementary Figure S3.

Stratified analysis of the predictive value of mean diameter for response to cTACE therapy based on tumor blood supply. ROC curve analysis for mean AER (A) and mean PER (B); C Hierarchical logistic regression analysis of mean diameter for response to cTACE; D Individual-level SHAP force plot of a patient from the external validation cohort with a highly vascularized but non-responsive HCC tumor. (TIFF 1871 kb)

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Zhang, L., Jin, Z., Li, C. et al. An interpretable machine learning model based on contrast-enhanced CT parameters for predicting treatment response to conventional transarterial chemoembolization in patients with hepatocellular carcinoma. Radiol med 129, 353–367 (2024). https://doi.org/10.1007/s11547-024-01785-z

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11547-024-01785-z

Keywords

Search

Navigation