An imaging-based artificial intelligence model for non-invasive grading of hepatic venous pressure gradient in cirrhotic portal hypertension

Qian Yu; Yifei Huang; Xiaoguo Li; Michael Pavlides; Dengxiang Liu; Hongwu Luo; Huiguo Ding; Weimin An; Fuquan Liu; Changzeng Zuo; Chunqiang Lu; Tianyu Tang; Yuancheng Wang; Shan Huang; Chuan Liu; Tianlei Zheng; Ning Kang; Changchun Liu; Jitao Wang; Seray Akçalar; Emrecan Çelebioğlu; Evren Üstüner; Sadık Bilgiç; Qu Fang; Chi Cheng Fu; Ruiping Zhang; Chengyan Wang; Jingwei Wei; Jie Tian; Necati Örmeci; Zeynep Ellik; Özgün Ömer Asiller; Shenghong Ju; Xiaolong Qi

doi:10.1016/j.xcrm.2022.100563

An imaging-based artificial intelligence model for non-invasive grading of hepatic venous pressure gradient in cirrhotic portal hypertension

Qian Yu
, Yifei Huang
, Xiaoguo Li
, Michael Pavlides
, Dengxiang Liu
, Hongwu Luo
, Huiguo Ding
, Weimin An
, Fuquan Liu
, Changzeng Zuo
, Chunqiang Lu
, Tianyu Tang
, Yuancheng Wang
, Shan Huang
, Chuan Liu
, Tianlei Zheng
, Ning Kang
, Changchun Liu
, Jitao Wang
, Seray Akçalar

Emrecan Çelebioğlu, Evren Üstüner, Sadık Bilgiç, Qu Fang, Chi Cheng Fu, Ruiping Zhang, Chengyan Wang, Jingwei Wei, Jie Tian, Necati Örmeci, Zeynep Ellik, Özgün Ömer Asiller, Shenghong Ju^*, Xiaolong Qi^*

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

39 Scopus citations

Abstract

The hepatic venous pressure gradient (HVPG) is the gold standard for cirrhotic portal hypertension (PHT), but it is invasive and specialized. Alternative non-invasive techniques are needed to assess the hepatic venous pressure gradient (HVPG). Here, we develop an auto-machine-learning CT radiomics HVPG quantitative model (aHVPG), and then we validate the model in internal and external test datasets by the area under the receiver operating characteristic curves (AUCs) for HVPG stages (≥10, ≥12, ≥16, and ≥20 mm Hg) and compare the model with imaging- and serum-based tools. The final aHVPG model achieves AUCs over 0.80 and outperforms other non-invasive tools for assessing HVPG. The model shows performance improvement in identifying the severity of PHT, which may help non-invasive HVPG primary prophylaxis when transjugular HVPG measurements are not available.

Original language	English
Article number	100563
Journal	Cell Reports Medicine
Volume	3
Issue number	3
DOIs	https://doi.org/10.1016/j.xcrm.2022.100563
State	Published - 15 Mar 2022
Externally published	Yes

Keywords

CHESS
CT
HVPG
cirrhosis
deep learning
machine learning
noninvasive tool
portal hypertension

Access to Document

10.1016/j.xcrm.2022.100563

Cite this

Yu, Q., Huang, Y., Li, X., Pavlides, M., Liu, D., Luo, H., Ding, H., An, W., Liu, F., Zuo, C., Lu, C., Tang, T., Wang, Y., Huang, S., Liu, C., Zheng, T., Kang, N., Liu, C., Wang, J., ... Qi, X. (2022). An imaging-based artificial intelligence model for non-invasive grading of hepatic venous pressure gradient in cirrhotic portal hypertension. Cell Reports Medicine, 3(3), Article 100563. https://doi.org/10.1016/j.xcrm.2022.100563

@article{86d7578b6fec453285a06dd477f40688,

title = "An imaging-based artificial intelligence model for non-invasive grading of hepatic venous pressure gradient in cirrhotic portal hypertension",

abstract = "The hepatic venous pressure gradient (HVPG) is the gold standard for cirrhotic portal hypertension (PHT), but it is invasive and specialized. Alternative non-invasive techniques are needed to assess the hepatic venous pressure gradient (HVPG). Here, we develop an auto-machine-learning CT radiomics HVPG quantitative model (aHVPG), and then we validate the model in internal and external test datasets by the area under the receiver operating characteristic curves (AUCs) for HVPG stages (≥10, ≥12, ≥16, and ≥20 mm Hg) and compare the model with imaging- and serum-based tools. The final aHVPG model achieves AUCs over 0.80 and outperforms other non-invasive tools for assessing HVPG. The model shows performance improvement in identifying the severity of PHT, which may help non-invasive HVPG primary prophylaxis when transjugular HVPG measurements are not available.",

keywords = "CHESS, CT, HVPG, cirrhosis, deep learning, machine learning, noninvasive tool, portal hypertension",

author = "Qian Yu and Yifei Huang and Xiaoguo Li and Michael Pavlides and Dengxiang Liu and Hongwu Luo and Huiguo Ding and Weimin An and Fuquan Liu and Changzeng Zuo and Chunqiang Lu and Tianyu Tang and Yuancheng Wang and Shan Huang and Chuan Liu and Tianlei Zheng and Ning Kang and Changchun Liu and Jitao Wang and Seray Ak{\c c}alar and Emrecan {\c C}elebioğlu and Evren {\"U}st{\"u}ner and Sadık Bilgi{\c c} and Qu Fang and Fu, \{Chi Cheng\} and Ruiping Zhang and Chengyan Wang and Jingwei Wei and Jie Tian and Necati {\"O}rmeci and Zeynep Ellik and Asiller, \{{\"O}zg{\"u}n {\"O}mer\} and Shenghong Ju and Xiaolong Qi",

note = "Publisher Copyright: {\textcopyright} 2022 The Authors",

year = "2022",

month = mar,

day = "15",

doi = "10.1016/j.xcrm.2022.100563",

language = "英语",

volume = "3",

journal = "Cell Reports Medicine",

issn = "2666-3791",

publisher = "Cell Press",

number = "3",

}

Yu, Q, Huang, Y, Li, X, Pavlides, M, Liu, D, Luo, H, Ding, H, An, W, Liu, F, Zuo, C, Lu, C, Tang, T, Wang, Y, Huang, S, Liu, C, Zheng, T, Kang, N, Liu, C, Wang, J, Akçalar, S, Çelebioğlu, E, Üstüner, E, Bilgiç, S, Fang, Q, Fu, CC, Zhang, R, Wang, C, Wei, J, Tian, J, Örmeci, N, Ellik, Z, Asiller, ÖÖ, Ju, S & Qi, X 2022, 'An imaging-based artificial intelligence model for non-invasive grading of hepatic venous pressure gradient in cirrhotic portal hypertension', Cell Reports Medicine, vol. 3, no. 3, 100563. https://doi.org/10.1016/j.xcrm.2022.100563

TY - JOUR

T1 - An imaging-based artificial intelligence model for non-invasive grading of hepatic venous pressure gradient in cirrhotic portal hypertension

AU - Yu, Qian

AU - Huang, Yifei

AU - Li, Xiaoguo

AU - Pavlides, Michael

AU - Liu, Dengxiang

AU - Luo, Hongwu

AU - Ding, Huiguo

AU - An, Weimin

AU - Liu, Fuquan

AU - Zuo, Changzeng

AU - Lu, Chunqiang

AU - Tang, Tianyu

AU - Wang, Yuancheng

AU - Huang, Shan

AU - Liu, Chuan

AU - Zheng, Tianlei

AU - Kang, Ning

AU - Liu, Changchun

AU - Wang, Jitao

AU - Akçalar, Seray

AU - Çelebioğlu, Emrecan

AU - Üstüner, Evren

AU - Bilgiç, Sadık

AU - Fang, Qu

AU - Fu, Chi Cheng

AU - Zhang, Ruiping

AU - Wang, Chengyan

AU - Wei, Jingwei

AU - Tian, Jie

AU - Örmeci, Necati

AU - Ellik, Zeynep

AU - Asiller, Özgün Ömer

AU - Ju, Shenghong

AU - Qi, Xiaolong

PY - 2022/3/15

Y1 - 2022/3/15

N2 - The hepatic venous pressure gradient (HVPG) is the gold standard for cirrhotic portal hypertension (PHT), but it is invasive and specialized. Alternative non-invasive techniques are needed to assess the hepatic venous pressure gradient (HVPG). Here, we develop an auto-machine-learning CT radiomics HVPG quantitative model (aHVPG), and then we validate the model in internal and external test datasets by the area under the receiver operating characteristic curves (AUCs) for HVPG stages (≥10, ≥12, ≥16, and ≥20 mm Hg) and compare the model with imaging- and serum-based tools. The final aHVPG model achieves AUCs over 0.80 and outperforms other non-invasive tools for assessing HVPG. The model shows performance improvement in identifying the severity of PHT, which may help non-invasive HVPG primary prophylaxis when transjugular HVPG measurements are not available.

AB - The hepatic venous pressure gradient (HVPG) is the gold standard for cirrhotic portal hypertension (PHT), but it is invasive and specialized. Alternative non-invasive techniques are needed to assess the hepatic venous pressure gradient (HVPG). Here, we develop an auto-machine-learning CT radiomics HVPG quantitative model (aHVPG), and then we validate the model in internal and external test datasets by the area under the receiver operating characteristic curves (AUCs) for HVPG stages (≥10, ≥12, ≥16, and ≥20 mm Hg) and compare the model with imaging- and serum-based tools. The final aHVPG model achieves AUCs over 0.80 and outperforms other non-invasive tools for assessing HVPG. The model shows performance improvement in identifying the severity of PHT, which may help non-invasive HVPG primary prophylaxis when transjugular HVPG measurements are not available.

KW - CHESS

KW - CT

KW - HVPG

KW - cirrhosis

KW - deep learning

KW - machine learning

KW - noninvasive tool

KW - portal hypertension

UR - https://www.scopus.com/pages/publications/85126313866

U2 - 10.1016/j.xcrm.2022.100563

DO - 10.1016/j.xcrm.2022.100563

M3 - 文章

C2 - 35492878

AN - SCOPUS:85126313866

SN - 2666-3791

VL - 3

JO - Cell Reports Medicine

JF - Cell Reports Medicine

IS - 3

M1 - 100563

ER -

An imaging-based artificial intelligence model for non-invasive grading of hepatic venous pressure gradient in cirrhotic portal hypertension

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this