Dataset Information

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.

SUBMITTER: Yu Q

PROVIDER: S-EPMC9040173 | biostudies-literature | 2022 Mar

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

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

Yu Qian Q Huang Yifei Y Li Xiaoguo X Pavlides Michael M Liu Dengxiang D Luo Hongwu H Ding Huiguo H An Weimin W Liu Fuquan F Zuo Changzeng C Lu Chunqiang C Tang Tianyu T Wang Yuancheng Y Huang Shan S Liu Chuan C Zheng Tianlei T Kang Ning N Liu Changchun C Wang Jitao J Akçalar Seray S Çelebioğlu Emrecan E Üstüner Evren E Bilgiç Sadık S Fang Qu Q Fu Chi-Cheng CC Zhang Ruiping R Wang Chengyan C Wei Jingwei J Tian Jie J Örmeci Necati N Ellik Zeynep Z Asiller Özgün Ömer ÖÖ Ju Shenghong S Qi Xiaolong X

Cell reports. Medicine 20220315 3

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 (<i>a</i>HVPG), 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, ...[more]

PMID: 35492878

Similar Datasets

Project description:BackgroundEsophagogastric variceal bleeding is a serious complication of decompensated cirrhosis. Transjugular intrahepatic portal shunt (TIPS) is a salvage treatment with clear hemostatic results. However, various complications may occur after TIPS, including postoperative liver failure, and the prognosis is very poor once occurs. Liver failure is a common symptom of severe liver disease with a high mortality rate. This study investigated the incidence of liver failure after TIPS treatment for varicose bleeding.MethodsWe analyzed the data of patients admitted to the First Affiliated Hospital of Soochow University between January 2013 and December 2018 with portal hypertension with an episode of acute gastroesophageal variceal bleeding. A total of 121 patients were referred to the regional liver unit for TIPS. Hepatic venous pressure gradient (HVPG) and clinical data were collected. Patients with incomplete data were excluded, and 93 patients were ultimately enrolled in the study. Primary outcomes were morbidity and hospital mortality within 4 weeks of surgery. The data were retrospectively and consecutively collected and evaluated by univariate and multivariate analyses to identify risk factors of liver failure.ResultsThe patients included 58 males (62.37%) and 35 females (37.63%), and the mean age was 58.43±11.85 years. The main cause was hepatitis B virus (HBV), which was found in 50.54% of patient. The overall surgical success rate was 83.87% (78/93). Of 15 treatment-failure patients, 9 (9.68%) died in hospital. Four patients died of liver failure, accounting for 44.44% of postoperative all-cause deaths. Univariate logistic regression analysis showed that only hepatic venous pressure gradient (HVPG) was an independent risk factor for post-TIPS morbidity [relative risk (RR) 1.156; 95% confidence interval (CI): 1.041 to 1.283; P=0.006]. In addition, HVPG was an independent risk factor for hospital mortality within 4 weeks (RR 1.133; 95% CI: 1.021 to 0.539; P=0.016).ConclusionsPost-TIPS liver failure is a serious complication in patients with cirrhosis. Pre-TIPS HVPG level may be used as a predictor of potential short-term postoperative adverse events.

Project description:Background and aimThe nuclear farnesoid X receptor (FXR) regulates critical pathways of hepatic metabolism, inflammation, and gut mucosal barrier. Thus, we investigated the association of FXR-single nucleotide polymorphism (SNPs) with hepatic decompensation and liver-related mortality in patients with advanced chronic liver disease.MethodsTwo FXR-SNPs (rs56163822 G > T and rs35724 G > C) were genotyped in a cohort of 402 prospectively characterized patients with hepatic venous pressure gradient (HVPG) ≥ 6 mmHg.ResultsOnly 19 patients (4.7%) harbored a rs56163822 T-allele and had less pronounced liver disease as indicated by lower Child-Pugh score (CPS, 6 ± 1 vs 7 ± 2 points, P = 0.034) and higher albumin levels (38.9 ± 4.9 vs 35.9 ± 5.9 g/L, P = 0.026). In contrast, n = 267 (66.4%) patients harbored minor rs35724 allele (G/C or C/C) and had more advanced liver disease, as indicated by a higher model of end-stage liver disease (11 ± 4 vs 10 ± 3, P = 0.016), while other baseline characteristics were similar across FXR-SNP genotypes. In compensated CPS-A patients, the rs35724 minor allele was independently protective for the development of ascites (adjusted hazard ratio [aHR] = 0.411, 95% confidence interval (95% CI): 0.191-0.885; P = 0.023) and tended to reduce the risk of hepatic decompensation (aHR = 0.625, 95% CI: 0.374-1.044, P = 0.072) in multivariate analyses. Of note, transplant-free survival was longer in patients with rs35724 minor allele and HVPG ≥ 10 mmHg (at 5 years: 68.2% vs 55.8%, P = 0.047) and those with HVPG ≥ 16 mmHg (63.3% vs 44.0%, P = 0.021). After adjusting for established risk factors, the rs35724 minor allele was independently associated with reduced liver-related mortality in the overall cohort (aHR = 0.658, 95% CI: 0.434-0.998, P = 0.049), in compensated CPS-A patients (aHR = 0.488, 95% CI: 0.252-0.946, P = 0.034), in patients with HVPG ≥ 10 mmHg (aHR = 0.547, 95% CI: 0.346-0.864, P = 0.010), and in patients with HVPG ≥ 16 mmHg (aHR = 0.519, 95% CI: 0.307-0.878, P = 0.014).ConclusionThe FXR-SNP rs35724 was associated with a reduced risk for development of ascites and liver-related mortality in patients with advanced chronic liver disease.

Dataset Information

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

Publications

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

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets