Machine learning-based ultrasound radiomics for predicting TP53 mutation status in hepatocellular carcinoma

ObjectivesTo explore the utility of machine learning-based ultrasound radiomics for predicting TP53 gene mutation in hepatocellular carcinoma (HCC).Methods154 HCC patients with 182 lesions from 2019 to 2024 were reviewed retrospectively. All lesions were randomly split into the training set (n = 129...

Full description

Saved in:
Bibliographic Details
Main Authors: Didi Bu, Shaobo Duan, Shanshan Ren, Yujing Ma, Yuanyuan Liu, Yahong Li, Xiguo Cai, Lianzhong Zhang
Format: Article
Language:English
Published: Frontiers Media S.A. 2025-04-01
Series:Frontiers in Medicine
Subjects:
Online Access:https://www.frontiersin.org/articles/10.3389/fmed.2025.1565618/full
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850140504359960576
author Didi Bu
Shaobo Duan
Shanshan Ren
Yujing Ma
Yuanyuan Liu
Yahong Li
Xiguo Cai
Lianzhong Zhang
Lianzhong Zhang
Lianzhong Zhang
author_facet Didi Bu
Shaobo Duan
Shanshan Ren
Yujing Ma
Yuanyuan Liu
Yahong Li
Xiguo Cai
Lianzhong Zhang
Lianzhong Zhang
Lianzhong Zhang
author_sort Didi Bu
collection DOAJ
description ObjectivesTo explore the utility of machine learning-based ultrasound radiomics for predicting TP53 gene mutation in hepatocellular carcinoma (HCC).Methods154 HCC patients with 182 lesions from 2019 to 2024 were reviewed retrospectively. All lesions were randomly split into the training set (n = 129) and the test set (n = 53), and ultrasound radiomics features were extracted and selected. Extreme gradient boosting tree (XGBoost), decision tree (DT), random forest (RF), support vector machine (SVM), and logistic regression (LR) were used to construct the ultrasound radiomics models, the clinical models, and the combined models. The predictive performance of various models was evaluated by the area under the curve (AUC), accuracy, calibration curve, and decision curve analysis (DCA).ResultsAmong the 182 lesions, 102 were confirmed as mutant TP53 and 80 were confirmed as wild-type TP53. The ultrasound radiomics model obtained an AUC of 0.778 and an accuracy of 0.774 in the test set. The clinical model achieved an AUC of 0.761 and an accuracy of 0.710 in the test set. Notably, integrating clinical features with ultrasound radiomics further enhanced predictive performance. The XGBoost-based combined model exhibited the highest predictive performance among all models, achieving an AUC of 0.846 and an accuracy of 0.823 in the test set. The decision curve analysis and calibration curve revealed that the XGBoost-based combined model provided the highest clinical benefit and exhibited strong predictive consistency.ConclusionMachine learning-based ultrasound radiomics signatures accurately predict TP53 gene mutations in HCC. The XGBoost-based combined model, which combined ultrasound radiomics features with clinical features, showed the best performance and represented a promising noninvasive approach for screening TP53-mutated HCC.
format Article
id doaj-art-a5ec96f8fe264342aad2dea5ab7f6485
institution OA Journals
issn 2296-858X
language English
publishDate 2025-04-01
publisher Frontiers Media S.A.
record_format Article
series Frontiers in Medicine
spelling doaj-art-a5ec96f8fe264342aad2dea5ab7f64852025-08-20T02:29:46ZengFrontiers Media S.A.Frontiers in Medicine2296-858X2025-04-011210.3389/fmed.2025.15656181565618Machine learning-based ultrasound radiomics for predicting TP53 mutation status in hepatocellular carcinomaDidi Bu0Shaobo Duan1Shanshan Ren2Yujing Ma3Yuanyuan Liu4Yahong Li5Xiguo Cai6Lianzhong Zhang7Lianzhong Zhang8Lianzhong Zhang9Department of Ultrasound, Zhengzhou University People’s Hospital, Henan Provincial People’s Hospital, Zhengzhou University, Zhengzhou, ChinaDepartment of Health Management, Henan Provincial People’s Hospital, Zhengzhou, ChinaDepartment of Ultrasound, Henan Provincial People’s Hospital, Zhengzhou, ChinaDepartment of Ultrasound, Henan University People’s Hospital, Henan Provincial People’s Hospital, Zhengzhou, ChinaDepartment of Ultrasound, Zhengzhou University People’s Hospital, Henan Provincial People’s Hospital, Zhengzhou University, Zhengzhou, ChinaDepartment of Ultrasound, Zhengzhou University People’s Hospital, Henan Provincial People’s Hospital, Zhengzhou University, Zhengzhou, ChinaHenan Rehabilitation Clinical Medical Research Center, Henan Provincial People’s Hospital, Zhengzhou, ChinaDepartment of Ultrasound, Zhengzhou University People’s Hospital, Henan Provincial People’s Hospital, Zhengzhou University, Zhengzhou, ChinaHenan Rehabilitation Clinical Medical Research Center, Henan Provincial People’s Hospital, Zhengzhou, ChinaHenan Key Laboratory for Ultrasound Molecular Imaging and Artificial Intelligence Medicine, Henan Provincial People’s Hospital, Zhengzhou, ChinaObjectivesTo explore the utility of machine learning-based ultrasound radiomics for predicting TP53 gene mutation in hepatocellular carcinoma (HCC).Methods154 HCC patients with 182 lesions from 2019 to 2024 were reviewed retrospectively. All lesions were randomly split into the training set (n = 129) and the test set (n = 53), and ultrasound radiomics features were extracted and selected. Extreme gradient boosting tree (XGBoost), decision tree (DT), random forest (RF), support vector machine (SVM), and logistic regression (LR) were used to construct the ultrasound radiomics models, the clinical models, and the combined models. The predictive performance of various models was evaluated by the area under the curve (AUC), accuracy, calibration curve, and decision curve analysis (DCA).ResultsAmong the 182 lesions, 102 were confirmed as mutant TP53 and 80 were confirmed as wild-type TP53. The ultrasound radiomics model obtained an AUC of 0.778 and an accuracy of 0.774 in the test set. The clinical model achieved an AUC of 0.761 and an accuracy of 0.710 in the test set. Notably, integrating clinical features with ultrasound radiomics further enhanced predictive performance. The XGBoost-based combined model exhibited the highest predictive performance among all models, achieving an AUC of 0.846 and an accuracy of 0.823 in the test set. The decision curve analysis and calibration curve revealed that the XGBoost-based combined model provided the highest clinical benefit and exhibited strong predictive consistency.ConclusionMachine learning-based ultrasound radiomics signatures accurately predict TP53 gene mutations in HCC. The XGBoost-based combined model, which combined ultrasound radiomics features with clinical features, showed the best performance and represented a promising noninvasive approach for screening TP53-mutated HCC.https://www.frontiersin.org/articles/10.3389/fmed.2025.1565618/fullradiomicshepatocellular carcinomamachine learningTP53ultrasonography
spellingShingle Didi Bu
Shaobo Duan
Shanshan Ren
Yujing Ma
Yuanyuan Liu
Yahong Li
Xiguo Cai
Lianzhong Zhang
Lianzhong Zhang
Lianzhong Zhang
Machine learning-based ultrasound radiomics for predicting TP53 mutation status in hepatocellular carcinoma
Frontiers in Medicine
radiomics
hepatocellular carcinoma
machine learning
TP53
ultrasonography
title Machine learning-based ultrasound radiomics for predicting TP53 mutation status in hepatocellular carcinoma
title_full Machine learning-based ultrasound radiomics for predicting TP53 mutation status in hepatocellular carcinoma
title_fullStr Machine learning-based ultrasound radiomics for predicting TP53 mutation status in hepatocellular carcinoma
title_full_unstemmed Machine learning-based ultrasound radiomics for predicting TP53 mutation status in hepatocellular carcinoma
title_short Machine learning-based ultrasound radiomics for predicting TP53 mutation status in hepatocellular carcinoma
title_sort machine learning based ultrasound radiomics for predicting tp53 mutation status in hepatocellular carcinoma
topic radiomics
hepatocellular carcinoma
machine learning
TP53
ultrasonography
url https://www.frontiersin.org/articles/10.3389/fmed.2025.1565618/full
work_keys_str_mv AT didibu machinelearningbasedultrasoundradiomicsforpredictingtp53mutationstatusinhepatocellularcarcinoma
AT shaoboduan machinelearningbasedultrasoundradiomicsforpredictingtp53mutationstatusinhepatocellularcarcinoma
AT shanshanren machinelearningbasedultrasoundradiomicsforpredictingtp53mutationstatusinhepatocellularcarcinoma
AT yujingma machinelearningbasedultrasoundradiomicsforpredictingtp53mutationstatusinhepatocellularcarcinoma
AT yuanyuanliu machinelearningbasedultrasoundradiomicsforpredictingtp53mutationstatusinhepatocellularcarcinoma
AT yahongli machinelearningbasedultrasoundradiomicsforpredictingtp53mutationstatusinhepatocellularcarcinoma
AT xiguocai machinelearningbasedultrasoundradiomicsforpredictingtp53mutationstatusinhepatocellularcarcinoma
AT lianzhongzhang machinelearningbasedultrasoundradiomicsforpredictingtp53mutationstatusinhepatocellularcarcinoma
AT lianzhongzhang machinelearningbasedultrasoundradiomicsforpredictingtp53mutationstatusinhepatocellularcarcinoma
AT lianzhongzhang machinelearningbasedultrasoundradiomicsforpredictingtp53mutationstatusinhepatocellularcarcinoma