| Receiver operating characteristic (ROC) analysis of selected texture .features and international normalized ratio (INR) for predicting high-risk portal hypertension (PH). (A) The ROC curve calculated by texture features of log.sigma.3.0.mm.3D_firstorder_RobustMeanAbsoluteDeviation, wavelet.LLH_ngtdm_Busyness, wavelet.HLL_glrlm_RunLengthNonUniformity, wavelet.HLH_glcm_MC, and wavelet.LLL_glrlm_RunLengthNonUniformity. (B) The ROC curve of INR for predicting high-risk PH.

| Receiver operating characteristic (ROC) analysis of selected texture .features and international normalized ratio (INR) for predicting high-risk portal hypertension (PH). (A) The ROC curve calculated by texture features of log.sigma.3.0.mm.3D_firstorder_RobustMeanAbsoluteDeviation, wavelet.LLH_ngtdm_Busyness, wavelet.HLL_glrlm_RunLengthNonUniformity, wavelet.HLH_glcm_MC, and wavelet.LLL_glrlm_RunLengthNonUniformity. (B) The ROC curve of INR for predicting high-risk PH.

Source publication
Article
Full-text available
Objective Clinical evidence suggests that the risk stratification of portal hypertension (PH) plays a vital role in disease progression and patient outcomes. However, the gold standard for stratifying PH [portal vein pressure (PVP) measurement] is invasive and therefore not suitable for routine clinical practice. This study is aimed to stratify PH...

Contexts in source publication

Context 1
... = 0.016). Receiver operating characteristic analysis showed that the above texture features had moderate capabilities to distinguish between the high-and low-risk PH groups, of which the best performance was achieved by the spleen-derived feature of wavelet.LLH_ngtdm_Busyness, with an AUC of 0.72, an accuracy of 0.746, a specificity of 0.681, and a sensitivity of 0.791 when using a cutoff value of 0.517 (Table 3 and Figure 3). The clinical feature of INR also showed a moderate performance for stratifying PH, with an AUC of 0.629, an accuracy of 0.649, a specificity of 0.468, and a sensitivity of 0.776 when using a cutoff value of 0.528 (Figure 3). ...
Context 2
... operating characteristic analysis showed that the above texture features had moderate capabilities to distinguish between the high-and low-risk PH groups, of which the best performance was achieved by the spleen-derived feature of wavelet.LLH_ngtdm_Busyness, with an AUC of 0.72, an accuracy of 0.746, a specificity of 0.681, and a sensitivity of 0.791 when using a cutoff value of 0.517 (Table 3 and Figure 3). The clinical feature of INR also showed a moderate performance for stratifying PH, with an AUC of 0.629, an accuracy of 0.649, a specificity of 0.468, and a sensitivity of 0.776 when using a cutoff value of 0.528 (Figure 3). ...
Context 3
... a result, texture features from the liver might not be able to reflect the complex hemodynamic changes of severe PH and may not correlate well with severe PH. However, as a relatively isolated organ, the spleen may not be influenced as much as the liver by the collateral circulation in the late stage of PH (37, 38); thus, the spleen-derived features seem more stable and reliable. The findings of this study suggest the potential of splenic texture features for the prediction of high-risk PH; however, due to the lack of relevant literature regarding the stratification of HR, hazard ratio; CI, confidence interval; PFS, progression-free survival. ...

Similar publications

Article
Full-text available
Background: Acoustic structure quantification (ASQ) has been applied to evaluate liver histologic changes by analyzing the speckle pattern seen on B-mode ultrasound. We aimed to assess the severity of portal hypertension (PHT) through hepatic ultrasonography. Methods: Sixty patients diagnosed with PHT and underwent surgical treatment with portos...

Citations

... In summary, the ML algorithm could identify fibrous matrix features and correlate with HVPG measurements. In addition to morphology-based features, texture features and deep convolutional neural network-based imaging analysis also showed a potential performance in assessing PH (28,57). However, due to the end-to-end nature of the model, specific features that helped improve diagnostic performance could not be elucidated. ...
Article
Full-text available
Background Portal hypertension monitoring is important throughout the natural course of cirrhosis. Hepatic venous pressure gradient (HVPG), regarded as the golden standard, is limited by invasiveness and technical difficulties. Portal hypertension is increasingly being assessed non-invasively, and hematological indices, imaging data, and statistical or computational models are studied to surrogate HVPG. This paper discusses the existing non-invasive methods based on measurement principles and reviews the methodological developments in the last 20 years. Methods First, we used VOSviewer to learn the architecture of this field. The publications about the non-invasive assessment of portal hypertension were retrieved from the Web of Science Core Collection (WoSCC). VOSviewer 1.6.17.0 was used to analyze and visualize these publications, including the annual trend, the study hotspots, the significant articles, authors, journals, and organizations in this field. Next, according to the cluster analysis result of the keywords, we further retrieved and classified the related studies to discuss. Results A total of 1,088 articles or review articles about our topic were retrieved from WoSCC. From 2000 to 2022, the number of publications is generally growing. “World Journal of Gastroenterology” published the most articles ( n = 43), while “Journal of Hepatology” had the highest citations. “Liver fibrosis” published in 2005 was the most influential manuscript. Among the 20,558 cited references of 1,088 retrieved manuscripts, the most cited was a study on liver stiffness measurement from 2007. The highest-yielding country was the United States, followed by China and Italy. “Berzigotti, Annalisa” was the most prolific author and had the most cooperation partners. Four study directions emerged from the keyword clustering: (1) the evaluation based on fibrosis; (2) the evaluation based on hemodynamic factors; (3) the evaluation through elastography; and (4) the evaluation of variceal bleeding. Conclusion The non-invasive assessment of portal hypertension is mainly based on two principles: fibrosis and hemodynamics. Liver fibrosis is the major initiator of cirrhotic PH, while hemodynamic factors reflect secondary alteration of splanchnic blood flow. Blood tests, US (including DUS and CEUS), CT, and magnetic resonance imaging (MRI) support the non-invasive assessment of PH by providing both hemodynamic and fibrotic information. Elastography, mainly USE, is the most important method of PH monitoring.
Article
Full-text available
    Background Personalized medicine requires the integration and analysis of vast amounts of patient data to realize individualized care. With Surgomics, we aim to facilitate personalized therapy recommendations in surgery by integration of intraoperative surgical data and their analysis with machine learning methods to leverage the potential of this data in analogy to Radiomics and Genomics. Methods We defined Surgomics as the entirety of surgomic features that are process characteristics of a surgical procedure automatically derived from multimodal intraoperative data to quantify processes in the operating room. In a multidisciplinary team we discussed potential data sources like endoscopic videos, vital sign monitoring, medical devices and instruments and respective surgomic features. Subsequently, an online questionnaire was sent to experts from surgery and (computer) science at multiple centers for rating the features’ clinical relevance and technical feasibility. Results In total, 52 surgomic features were identified and assigned to eight feature categories. Based on the expert survey (n = 66 participants) the feature category with the highest clinical relevance as rated by surgeons was “surgical skill and quality of performance” for morbidity and mortality (9.0 ± 1.3 on a numerical rating scale from 1 to 10) as well as for long-term (oncological) outcome (8.2 ± 1.8). The feature category with the highest feasibility to be automatically extracted as rated by (computer) scientists was “Instrument” (8.5 ± 1.7). Among the surgomic features ranked as most relevant in their respective category were “intraoperative adverse events”, “action performed with instruments”, “vital sign monitoring”, and “difficulty of surgery”. Conclusion Surgomics is a promising concept for the analysis of intraoperative data. Surgomics may be used together with preoperative features from clinical data and Radiomics to predict postoperative morbidity, mortality and long-term outcome, as well as to provide tailored feedback for surgeons. Graphical abstract