CT-scan Based Liver and Spleen Volume Measurement as a Prognostic Indicator for Patients with Cirrhosis

AI model using CT-based imaging biomarkers to predict hepatocellular carcinoma in patients with chronic hepatitis B

BACKGROUND & AIMS

Various hepatocellular carcinoma (HCC) prediction models have been proposed for patients with chronic hepatitis B (CHB) using clinical variables. We aimed to develop an artificial intelligence (AI)-based HCC prediction model by incorporating imaging biomarkers derived from abdominal computed tomography (CT) images along with clinical variables.

METHODS

An AI prediction model employing a gradient-boosting machine algorithm was developed utilizing imaging biomarkers extracted by DeepFore, a deep learning-based CT auto-segmentation software. The derivation cohort (n = 5,585) was randomly divided into the training and internal validation sets at a 3:1 ratio. The external validation cohort included 2,883 patients. Six imaging biomarkers (i.e. abdominal visceral fat-total fat volume ratio, total fat-trunk volume ratio, spleen volume, liver volume, liver-spleen Hounsfield unit ratio, and muscle Hounsfield unit) and eight clinical variables were selected as the main variables of our model, PLAN-B-DF.

RESULTS

In the internal validation set (median follow-up duration = 7.4 years), PLAN-B-DF demonstrated an excellent predictive performance with a c-index of 0.91 and good calibration function (p = 0.78 by the Hosmer-Lemeshow test). In the external validation cohort (median follow-up duration = 4.6 years), PLAN-B-DF showed a significantly better discrimination function compared to previous models, including PLAN-B, PAGE-B, modified PAGE-B, and CU-HCC (c-index, 0.89 vs. 0.65-0.78; all p <0.001), and maintained a good calibration function (p = 0.42 by the Hosmer-Lemeshow test). When patients were classified into four groups according to the risk probability calculated by PLAN-B-DF, the 10-year cumulative HCC incidence was 0.0%, 0.4%, 16.0%, and 46.2% in the minimal-, low-, intermediate-, and high-risk groups, respectively.

CONCLUSION

This AI prediction model, integrating deep learning-based auto-segmentation of CT images, offers improved performance in predicting HCC risk among patients with CHB compared to previous models.

IMPACT AND IMPLICATIONS

The novel predictive model PLAN-B-DF, employing an automated computed tomography segmentation algorithm, significantly improves predictive accuracy and risk stratification for hepatocellular carcinoma in patients with chronic hepatitis B (CHB). Using a gradient-boosting algorithm and computed tomography metrics, such as visceral fat volume and myosteatosis, PLAN-B-DF outperforms previous models based solely on clinical and demographic data. This model not only shows a higher c-index compared to previous models, but also effectively classifies patients with CHB into different risk groups. This model uses machine learning to analyze the complex relationships among various risk factors contributing to hepatocellular carcinoma occurrence, thereby enabling more personalized surveillance for patients with CHB.

Application of ultrasound elastography and splenic size in predicting post-hepatectomy liver failure: Unveiling new clinical perspectives

In this article, we discuss the study by Cheng et al, published in the World Journal of Gastroenterology, focusing on predictive methods for post-hepatectomy liver failure (PHLF). PHLF is a common and serious complication, and accurate prediction is critical for clinical management. The study examines the potential of ultrasound elastography and splenic size in predicting PHLF. Ultrasound elastography reflects liver functional reserve, while splenic size provides additional predictive value. By integrating these factors with serological markers, we developed a comprehensive prediction model that effectively stratifies patient risk and supports personalized clinical decisions. This approach offers new insights into predicting PHLF. These methods not only assist clinicians in identifying high-risk patients earlier but also provide scientific support for personalized treatment strategies. Future research will aim to validate the model's accuracy with larger sample sizes, further enhancing the clinical application of these non-invasive indicators.

Predicting High-Risk Esophageal Varices in Cirrhosis: A Multi-Parameter Splenic CT Study

RATIONALE AND OBJECTIVES

To explore the value of splenic hemodynamic parameters from low-dose one-stop dual-energy and perfusion CT (LD-DE&PCT) in non-invasively predicting high-risk esophageal varices (HREV) in cirrhotic patients.

METHODS

We retrospectively analyzed cirrhotic patients diagnosed with esophageal varices (EV) through clinical, laboratory, imaging, and endoscopic examinations from September 2021 to December 2023 in our hospital. All patients underwent LD-DE&PCT to acquire splenic iodine concentration and perfusion parameters. Radiation dose was recorded. Patients were classified into non-HREV and HREV groups based on endoscopy. Univariate and multivariate logistic regression analysis were performed, and the prediction model for HREV was constructed. P < 0.05 was considered statistically significant.

RESULTS

Univariate analysis revealed that significant differences were found in portal iodine concentration (PIC), blood flow (BF), permeability surface (PS), spleen volume (V-S), total iodine concentration (TIC), and total blood volume of the spleen (BV-S) between groups. TIC demonstrated the highest predictive value with an area under the curve (AUC) value of 0.87. Multivariate analysis showed that PIC, PS, and BV-S were independent risk factors for HREV. The logistic regression model for predicting HREV had an AUC of 0.93. The total radiation dose was 20.66 ± 4.07 mSv.

CONCLUSION

Splenic hemodynamic parameters obtained from LD-DE&PCT can non-invasively and accurately assess the hemodynamic status of the spleen in cirrhotic patients with EV and predict the occurrence of HREV. Despite the retrospective study design and limited sample size of this study, these findings deserve further validation through prospective studies with larger cohorts.

Development of a non-invasive diagnostic model for high-risk esophageal varices based on radiomics of spleen CT

PURPOSE

To evaluate the diagnostic performance of radiomics models derived from multi-phase spleen CT for high-risk esophageal varices (HREV) in cirrhotic patients.

METHODS

We retrospectively selected cirrhotic patients with esophageal varices from two hospitals from September 2019 to September 2023. Patients underwent non-contrast and contrast-enhanced CT scans and were categorized into HREV and non-HREV groups based on endoscopic evaluations. Radiomics features were extracted from spleen CT images in non-contrast, arterial, and portal venous phases, with feature selection via lasso regression and Pearson's correlation. Ten machine learning models were developed to diagnose HREV, evaluated by area under the curve (AUC). The AUC values of the three groups of models were statistically compared by the Kruskal-Wallis H test and Bonferroni-corrected Mann-Whitney U test. A p-value less than 0.05 was considered statistically significant.

RESULTS

Among 233 patients, 11, 6, and 11 features were selected from non-contrast, arterial, and portal venous phases, respectively. Significant differences in AUC values were observed across phases (p < 0.05), and the arterial phase models showed the highest AUC values. The best model in arterial phase was the logical regression model, whose AUC value was 0.85, sensitivity was 83.3%, specificity was 80% and F1 score was 0.81.

CONCLUSION

Radiomics models based on spleen CT, especially the arterial phase models, demonstrate high diagnostic accuracy for HREV, offering the potential for early detection and intervention.

Spleen volume is associated with overt hepatic encephalopathy after transjugular intrahepatic portosystemic shunt in patients with portal hypertension

BACKGROUND

Portal shunt and immune status related to the spleen are related to the occurrence of hepatic encephalopathy (HE). It is unknown whether spleen volume before transjugular intrahepatic portosystemic shunt (TIPS) is related to postoperative HE.

AIM

To investigate the relationship between spleen volume and the occurrence of HE.

METHODS

This study included 135 patients with liver cirrhosis who underwent TIPS, and liver and spleen volumes were elevated upon computed tomography imaging. The Kaplan-Meier curve was used to compare the difference in the incidence rate of HE among patients with different spleen volumes. Univariate and multivariate Cox regression analyses were performed to identify the factors affecting overt HE (OHE). Restricted cubic spline was used to examine the shapes of the dose-response association between spleen volumes and OHE risk.

RESULTS

The results showed that 37 (27.2%) of 135 patients experienced OHE during a 1-year follow-up period. Compared with preoperative spleen volume (901.30 ± 471.90 cm3), there was a significant decrease in spleen volume after TIPS (697.60 ± 281.0 cm3) in OHE patients. As the severity of OHE increased, the spleen volume significantly decreased (P < 0.05). Compared with patients with a spleen volume ≥ 782.4 cm3, those with a spleen volume < 782.4 cm3 had a higher incidence of HE (P < 0.05). Cox regression analysis showed that spleen volume was an independent risk factor for post-TIPS OHE (hazard ratio = 0.494, P < 0.05). Restricted cubic spline model showed that with an increasing spleen volume, OHE risk showed an initial increase and then decrease (P < 0.05).

CONCLUSION

Spleen volume is related to the occurrence of OHE after TIPS. Preoperative spleen volume is an independent risk factor for post-TIPS OHE.

Nomogram based on liver stiffness and spleen area with ultrasound for posthepatectomy liver failure: A multicenter study

BACKGROUND

Liver stiffness (LS) measurement with two-dimensional shear wave elastography (2D-SWE) correlates with the degree of liver fibrosis and thus indirectly reflects liver function reserve. The size of the spleen increases due to tissue proliferation, fibrosis, and portal vein congestion, which can indirectly reflect the situation of liver fibrosis/cirrhosis. It was reported that the size of the spleen was related to posthepatectomy liver failure (PHLF). So far, there has been no study combining 2D-SWE measurements of LS with spleen size to predict PHLF. This prospective study aimed to investigate the utility of 2D-SWE assessing LS and spleen area (SPA) for the prediction of PHLF in hepatocellular carcinoma (HCC) patients and to develop a risk prediction model.

AIM

To investigate the utility of 2D-SWE assessing LS and SPA for the prediction of PHLF in HCC patients and to develop a risk prediction model.

METHODS

This was a multicenter observational study prospectively analyzing patients who underwent hepatectomy from October 2020 to March 2022. Within 1 wk before partial hepatectomy, ultrasound examination was performed to measure LS and SPA, and blood was drawn to evaluate the patient's liver function and other conditions. Least absolute shrinkage and selection operator logistic regression and multivariate logistic regression analysis was applied to identify independent predictors of PHLF and develop a nomogram. Nomogram performance was validated further. The diagnostic performance of the nomogram was evaluated with receiver operating characteristic curve compared with the conventional models, including the model for end-stage liver disease (MELD) score and the albumin-bilirubin (ALBI) score.

RESULTS

A total of 562 HCC patients undergoing hepatectomy (500 in the training cohort and 62 in the validation cohort) were enrolled in this study. The independent predictors of PHLF were LS, SPA, range of resection, blood loss, international normalized ratio, and total bilirubin. Better diagnostic performance of the nomogram was obtained in the training [area under receiver operating characteristic curve (AUC): 0.833; 95% confidence interval (95%CI): 0.792-0.873; sensitivity: 83.1%; specificity: 73.5%] and validation (AUC: 0.802; 95%CI: 0.684-0.920; sensitivity: 95.5%; specificity: 52.5%) cohorts compared with the MELD score and the ALBI score.

CONCLUSION

This PHLF nomogram, mainly based on LS by 2D-SWE and SPA, was useful in predicting PHLF in HCC patients and presented better than MELD score and ALBI score.

The impact of hepatic and splenic volumetric assessment in imaging for chronic liver disease: a narrative review

Chronic liver disease is responsible for significant morbidity and mortality worldwide. Abdominal computed tomography (CT) and magnetic resonance imaging (MRI) can fully visualise the liver and adjacent structures in the upper abdomen providing a reproducible assessment of the liver and biliary system and can detect features of portal hypertension. Subjective interpretation of CT and MRI in the assessment of liver parenchyma for early and advanced stages of fibrosis (pre-cirrhosis), as well as severity of portal hypertension, is limited. Quantitative and reproducible measurements of hepatic and splenic volumes have been shown to correlate with fibrosis staging, clinical outcomes, and mortality. In this review, we will explore the role of volumetric measurements in relation to diagnosis, assessment of severity and prediction of outcomes in chronic liver disease patients. We conclude that volumetric analysis of the liver and spleen can provide important information in such patients, has the potential to stratify patients' stage of hepatic fibrosis and disease severity, and can provide critical prognostic information. CRITICAL RELEVANCE STATEMENT: This review highlights the role of volumetric measurements of the liver and spleen using CT and MRI in relation to diagnosis, assessment of severity, and prediction of outcomes in chronic liver disease patients. KEY POINTS: Volumetry of the liver and spleen using CT and MRI correlates with hepatic fibrosis stages and cirrhosis. Volumetric measurements correlate with chronic liver disease outcomes. Fully automated methods for volumetry are required for implementation into routine clinical practice.

A nomogram for predicting prognosis in patients with transjugular intrahepatic portosystemic shunt creation based on deep learning-derived spleen volume-to-platelet ratio

OBJECTIVES

The objective of our study was to develop a nomogram to predict post-transjugular intrahepatic portosystemic shunt (TIPS) survival in patients with cirrhosis based on CT images.

METHODS

This retrospective cohort study included patients who had received TIPS operation at the Wenzhou Medical University First Affiliated Hospital between November 2013 and April 2017. To predict prognosis, a nomogram and Web-based probability were developed to assess the overall survival (OS) rates at 1, 3, and 5 years based on multivariate analyses. With deep learning algorithm, the automated measurement of liver and spleen volumes can be realized. We assessed the predictive accuracy and discriminative ability of the nomogram using the concordance index (C-index), receiver operating characteristic (ROC) curve, calibration curve, and decision curve analysis (DCA).

RESULTS

Age, total bilirubin, and spleen volume-to-platelet ratio (SVPR) were identified as the independent risk factors for OS. The nomogram was constructed based on the above risk factors. The C-index (0.80, 0.74, 0.70), ROC curve (area under curve: 0.828, 0.761, 0.729), calibration curve, and DCA showed that nomogram good at predictive value, stability, and clinical benefit in the prediction of 1-, 3-, 5-year OS in patients with TIPS creation.

CONCLUSIONS

We constructed a nomogram for predicting prognosis in patients with TIPS creation based on risk factors. The nomogram can help clinicians in identifying patients with poor prognosis, eventually facilitating earlier treatment and selecting suitable patients before TIPS.

ADVANCES IN KNOWLEDGE

This study developed the first nomogram based on SVPR to predict the prognosis of patients treated with TIPS. The nomogram could help clinician in non-invasive decision-making.

Opportunistic Extraction of Quantitative CT Biomarkers: Turning the Incidental Into Prognostic Information

The past two decades have seen a significant increase in the use of CT, with a corresponding rise in the mean population radiation dose. This rise in CT use has caused improved diagnostic certainty in conditions that were not previously routinely evaluated using CT, such as headaches, back pain, and chest pain. Unused data, unrelated to the primary diagnosis, embedded within these scans have the potential to provide organ-specific measurements that can be used to prognosticate or risk-profile patients for a wide variety of conditions. The recent increased availability of computing power, expertise and software for automated segmentation and measurements, assisted by artificial intelligence, provides a conducive environment for the deployment of these analyses into routine use. Data gathering from CT has the potential to add value to examinations and help offset the public perception of harm from radiation exposure. We review the potential for the collection of these data and propose the incorporation of this strategy into routine clinical practice.

Deep Learning for Computed Tomography Assessment of Hepatic Fibrosis and Cirrhosis: A Systematic Review

Studies were identified using deep learning artificial intelligence methods for the analysis of computed tomography images in the assessment of hepatic fibrosis and cirrhosis. A systematic review was conducted in line with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses of Diagnostic Test Accuracy Studies protocol to evaluate the accuracy of deep learning algorithms for this objective (PROSPERO CRD 42023366201). A literature search was conducted on Medline, Embase, Web of Science, and IEEE Xplore databases. The search was conducted with a timeline from January 1, 2000,through November 13, 2022. Our search resulted in 3877 studies for screening, which yielded 6 studies meeting our inclusion criteria. All studies were retrospective. Three studies performed internal validation, and 2 studies performed external validation. Four studies used image classification algorithms, whereas 2 studies used image segmentation algorithms to derive volumetric measurements of the liver and spleen. Accuracy of the algorithms was variable in diagnosing significant and advanced fibrosis and cirrhosis, with the area under the curve ranging from 0.63 to 0.97. Deep learning algorithms using computed tomography images have the potential to classify fibrosis stages. Heterogeneity in cohorts and methodologies limits the generalizability of these studies.

Non-invasive model for predicting high-risk esophageal varices based on liver and spleen stiffness

BACKGROUND

Acute bleeding due to esophageal varices (EVs) is a life-threatening complication in patients with cirrhosis. The diagnosis of EVs is mainly through upper gastrointestinal endoscopy, but the discomfort, contraindications and complications of gastrointestinal endoscopic screening reduce patient compliance. According to the bleeding risk of EVs, the Baveno VI consensus divides varices into high bleeding risk EVs (HEVs) and low bleeding risk EVs (LEVs). We sought to identify a non-invasive prediction model based on spleen stiffness measurement (SSM) and liver stiffness measurement (LSM) as an alternative to EVs screening.

AIM

To develop a safe, simple and non-invasive model to predict HEVs in patients with viral cirrhosis and identify patients who can be exempted from upper gastrointestinal endoscopy.

METHODS

Data from 200 patients with viral cirrhosis were included in this study, with 140 patients as the modelling group and 60 patients as the external validation group, and the EVs types of patients were determined by upper gastrointestinal endoscopy and the Baveno VI consensus. Those patients were divided into the HEVs group (66 patients) and the LEVs group (74 patients). The effect of each parameter on HEVs was analyzed by univariate and multivariate analyses, and a non-invasive prediction model was established. Finally, the discrimination ability, calibration ability and clinical efficacy of the new model were verified in the modelling group and the external validation group.

RESULTS

Univariate and multivariate analyses showed that SSM and LSM were associated with the occurrence of HEVs in patients with viral cirrhosis. On this basis, logistic regression analysis was used to construct a prediction model: Ln [P/(1-P)] = -8.184 -0.228 × SSM + 0.642 × LSM. The area under the curve of the new model was 0.965. When the cut-off value was 0.27, the sensitivity, specificity, positive predictive value and negative predictive value of the model for predicting HEVs were 100.00%, 82.43%, 83.52%, and 100%, respectively. Compared with the four prediction models of liver stiffness-spleen diameter to platelet ratio score, variceal risk index, aspartate aminotransferase to alanine aminotransferase ratio, and Baveno VI, the established model can better predict HEVs in patients with viral cirrhosis.

CONCLUSION

Based on the SSM and LSM measured by transient elastography, we established a non-invasive prediction model for HEVs. The new model is reliable in predicting HEVs and can be used as an alternative to routine upper gastrointestinal endoscopy screening, which is helpful for clinical decision making.

Nomogram based on spleen volume expansion rate predicts esophagogastric varices bleeding risk in patients with hepatitis B liver cirrhosis

BACKGROUND

We aimed to explore the risk factors for hemorrhage of esophagogastric varices (EGVs) in patients with hepatitis B cirrhosis and to construct a novel nomogram model based on the spleen volume expansion rate to predict the risk of esophagogastric varices bleeding.

METHODS

Univariate and multivariate logistic regression analysis was used to analyze the risk factors for EGVs bleeding. Nomograms were established based on the multivariate analysis results. The predictive accuracy of the nomograms was assessed using the area under the curve (AUC or C-index) of the receiver operating characteristic (ROC) and calibration curves. Decision curve analysis was used to determine the clinical benefit of the nomogram. We created a nomogram of the best predictive models.

RESULTS

A total of 142 patients' hepatitis B cirrhosis with esophagogastric varices were included in this study, of whom 85 (59.9%) had a history of EGVs bleeding and 57 (40.1%) had no EGVs bleeding. The spleen volume expansion rate, serum sodium levels (mmol/L), hemoglobin levels (g/L), and prothrombin time (s) were independent predictors for EGVs bleeding in patients with hepatitis B liver cirrhosis (P < 0.05). The above predictors were included in the nomogram prediction model. The area under the ROC curve (AUROC) of the nomogram was 0.781, the C-index obtained by internal validation was 0.757, and the calibration prediction curve fit well with the ideal curve. The AUROCs of the PLT-MELD and APRI were 0.648 and 0.548, respectively.

CONCLUSION

In this study, a novel nomogram for predicting the risk of EGVs bleeding in patients with hepatitis B cirrhosis was successfully constructed by combining the spleen volume expansion rate, serum sodium levels, hemoglobin levels, and prothrombin time. The predictive model can provide clinicians with a reference to help them make clinical decisions.