A Machine Learning Model Based on Counterfactual Theory for Treatment Decision of Hepatocellular Carcinoma Patients - Original Research

Q: What treatment options does the model evaluate for hepatocellular carcinoma patients?

The model evaluates hepatectomy and transarterial chemoembolization (TACE) and predicts outcome differences between these two treatment modalities.

Q: How are causal effects and bias handled in the study design?

Propensity score matching is used to obtain perfectly matched patients for the independent test cohort, and inverse probability of treatment weighting controls bias for unmatched patients used in the training cohort.

Q: What inputs are used to build the machine learning models?

The models combine clinical characteristics selected through Cox proportional hazards regression with radiomics features selected using correlation analysis and random survival forest.

Aug 2, 2026

Upload by Aria

A machine learning framework based on counterfactual theory is developed to predict treatment efficacy in hepatocellular carcinoma patients receiving hepatectomy or transarterial chemoembolization (TACE). A retrospective cohort of patients treated between June 2016 and July 2021 is analyzed using propensity score matching and inverse probability of treatment weighting to form independent test and training cohorts. Clinical variables and radiomics features are selected via survival regression and feature selection methods, then models estimate post-treatment death probabilities and recommend optimal regimens, while a prognostic model predicts overall survival.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages13

Reading Time33 min

AccessFree to download

Introduction

Treatment guidelines and decision challenges

Uncertainty between hepatectomy and TACE

Purpose

Patients and Methods

Results

Conclusion

Journal of Hepatocellular Carcinoma downloaded from [https://www.dovepress.com/](https://www.dovepress.com/) For personal use only. Journal of Hepatocellular Carcinoma Dovepress open access to scientific and medical research Open Access Full Text Article ORIGINAL RESEARCH A Machine Learning Model Based on Counterfactual Theory for Treatment Decision of Hepatocellular Carcinoma Patients Xiaoqin Wei 1 , *, Fang Wang2 , *, Ying Liu 3 , Zeyong Li 4 , Zhong Xue2 , Mingyue Tang 5 , Xiaowen Chen 1 1School of Medical Imaging, North Sichuan Medical College, Nanchong City, Sichuan Province, People’s Republic of China; 2Department of Research and Development, Shanghai United Imaging Intelligence Co., Ltd, Shanghai, People’s Republic of China; 3Department of Radiology, The First Affiliated Hospital of Chengdu Medical College, Chengdu City, Sichuan Province, People’s Republic of China; 4Department of Radiology, Bishan Hospital of Chongqing Medical University, ChongQing, People’s Republic of China; 5Department of Physics, School of Basic Medicine, North Sichuan Medical College, Nanchong, People’s Republic of China *These authors contributed equally to this work Correspondence: Xiaowen Chen, School of Medical Imaging, North Sichuan Medical College, 234 Fujiang Road, Nanchong City, Sichuan Province, People’s Republic of China, 637001, Email [56833804@qq.com](56833804@qq.com) | Purpose: To predict the efficacy of patients treated with hepatectomy and transarterial chemoembolization (TACE) based on machine learning models using clinical and radiomics features.
Patients and Methods: Patients with HCC whose first treatment was hepatectomy or TACE from June 2016 to July 2021 were collected in the retrospective cohort study. To ensure a causal effect of treatment effect and treatment modality, perfectly matched patients were obtained according to the principle of propensity score matching and used as an independent test cohort. Inverse probability of treatment weighting was used to control bias for unmatched patients, and the weighted results were used as the training cohort. Clinical characteristics were selected by univariate and multivariate analysis of cox proportional hazards regression, and radiomics features were selected using correlation analysis and random survival forest. The machine learning models (Deathhepatectomy and DeathTACE) were constructed to predict the probability of patient death after treatment (hepatectomy and TACE) by combining clinical and radiomics features, and an optimal treatment regimen was recommended. In addition, a prognostic model was constructed to predict the survival time of all patients.
Results: A total of 418 patients with HCC who received either hepatectomy (n=267, mean age, 58 years ± 11 [standard deviation]; 228 men) or TACE (n=151, mean age, 59 years ± 13 [standard deviation]; 127 men) were recruited. After constructing the machine learning models Deathhepatectomy and DeathTACE, patients were divided into the hepatectomy-preferred and TACE-preferred groups. In the hepatectomy-preferred group, hepatectomy had a significantly prolonged survival time than TACE (training cohort: P < 0.001; testing cohort: P < 0.001), and vise versa for the TACE-preferred group. In addition, the prognostic model yielded high predictive capability for overall survival.
Conclusion: The machine learning models could predict the outcomes difference between hepatectomy and TACE, and prognostic models could predict the overall survival for HCC patients.
Keywords: radiomics, hepatocellular carcinoma, prognosis, hepatectomy | | --- | | Introduction
Different treatment guidelines were proposed to provide appropriate treatment options for hepatocellular carcinoma (HCC) patients, such as China Liver Cancer (CNLC) staging 1 and Barcelona Clinic Liver Cancer (BCLC) staging.2 However, they might be inadequate for clinical decision-making. First, there is disagreement among treatment guidelines over the

FAQs

What treatment options does the model evaluate for hepatocellular carcinoma patients?

The model evaluates hepatectomy and transarterial chemoembolization (TACE) and predicts outcome differences between these two treatment modalities.

How are causal effects and bias handled in the study design?

Propensity score matching is used to obtain perfectly matched patients for the independent test cohort, and inverse probability of treatment weighting controls bias for unmatched patients used in the training cohort.

What inputs are used to build the machine learning models?

The models combine clinical characteristics selected through Cox proportional hazards regression with radiomics features selected using correlation analysis and random survival forest.

Aug 2, 2026

Upload by Aria