Prediction and Diagnosis of Non-Alcoholic Fatty Liver Disease (NAFLD) and Identification of Its Associated Factors Using the Classification Tree Method

Application of artificial intelligence techniques for non-alcoholic fatty liver disease diagnosis: A systematic review (2005-2023)

BACKGROUND AND OBJECTIVES

Non-alcoholic fatty liver disease (NAFLD) is a common liver disease with a rapidly growing incidence worldwide. For prognostication and therapeutic decisions, it is important to distinguish the pathological stages of NAFLD: steatosis, steatohepatitis, and liver fibrosis, which are definitively diagnosed on invasive biopsy. Non-invasive ultrasound (US) imaging, including US elastography technique, and clinical parameters can be used to diagnose and grade NAFLD and its complications. Artificial intelligence (AI) is increasingly being harnessed for developing NAFLD diagnostic models based on clinical, biomarker, or imaging data. In this work, we systemically reviewed the literature for AI-enabled NAFLD diagnostic models based on US (including elastography) and clinical (including serological) data.

METHODS

We performed a comprehensive search on Google Scholar, Scopus, and PubMed search engines for articles published between January 2005 and June 2023 related to AI models for NAFLD diagnosis based on US and/or clinical parameters using the following search terms: "non-alcoholic fatty liver disease", "non-alcoholic steatohepatitis", "deep learning", "machine learning", "artificial intelligence", "ultrasound imaging", "sonography", "clinical information".

RESULTS

We reviewed 64 published models that used either US (including elastography) or clinical data input to detect the presence of NAFLD, non-alcoholic steatohepatitis, and/or fibrosis, and in some cases, the severity of steatosis, inflammation, and/or fibrosis as well. The performances of the published models were summarized, and stratified by data input and algorithms used, which could be broadly divided into machine and deep learning approaches.

CONCLUSION

AI models based on US imaging and clinical data can reliably detect NAFLD and its complications, thereby reducing diagnostic costs and the need for invasive liver biopsy. The models offer advantages of efficiency, accuracy, and accessibility, and serve as virtual assistants for specialists to accelerate disease diagnosis and reduce treatment costs for patients and healthcare systems.

Prevalence of non-alcoholic fatty liver and its related factors in Iran: Systematic review and meta-analysis

Non-alcoholic fatty liver disease (NAFLD) is a systemic disorder with a complex multifactorial and heterogeneous pathogenesis and has become the most common cause of chronic liver disease in many countries around the world. Numerous studies in Iran have presented different results on the prevalence and risk factors of NAFLD, in this study, which has been done in a systematic review and meta-analysis, provides a good estimate of the prevalence and risk factors of the disease in Iran. Following the peer review of electronic search strategies (PRESS and the preferred reporting items for systematic reviews and meta-analyses [PRISMA] statement, we searched Web of Science, PubMed, Embase, Scopus, and Persian scientific searcher (Elmnet) from inception to September 19, 2022. In the present study, 71 articles were reviewed for qualitative and meta-analysis. The overall mean prevalence of NAFLD in children studies was 22.4% (95% confidence interval [CI]: 10.9% to 33.9%). The prevalence was notably higher in adult studies 40.5% (95% CI: 35.1% to 46%). In 24 studies, the association between NAFLD and sex was reported, 10 of which showed significant relationships. Out of 46 studies observed that NAFLD prevalence increased significantly with body mass index (BMI). Eight out of 14 studies reported significant associations between FBS and NAFLD in children's studies. Though Iran has a high NAFLD prevalence compared to most areas, and due to the unfavorable situation of risk factors contributing to the NAFLD, it is necessary to take the necessary interventions to control these risk factors and prevent NAFLD.

Liver Fat Scores for Noninvasive Diagnosis and Monitoring of Nonalcoholic Fatty Liver Disease in Epidemiological and Clinical Studies

Nonalcoholic fatty liver disease (NAFLD) is strongly associated with the metabolic syndrome and type 2 diabetes and independently contributes to long-term complications. Being often asymptomatic but reversible, it would require population-wide screening, but direct diagnostics are either too invasive (liver biopsy), costly (MRI) or depending on the examiner's expertise (ultrasonography). Hepatosteatosis is usually accommodated by features of the metabolic syndrome (e.g. obesity, disturbances in triglyceride and glucose metabolism), and signs of hepatocellular damage, all of which are reflected by biomarkers, which poorly predict NAFLD as single item, but provide a cheap diagnostic alternative when integrated into composite liver fat indices. Fatty liver index, NAFLD LFS, and hepatic steatosis index are common and accurate indices for NAFLD prediction, but show limited accuracy for liver fat quantification. Other indices are rarely used. Hepatic fibrosis scores are commonly used in clinical practice, but their mandatory reflection of fibrotic reorganization, hepatic injury or systemic sequelae reduces sensitivity for the diagnosis of simple steatosis. Diet-induced liver fat changes are poorly reflected by liver fat indices, depending on the intervention and its specific impact of weight loss on NAFLD. This limited validity in longitudinal settings stimulates research for new equations. Adipokines, hepatokines, markers of cellular integrity, genetic variants but also simple and inexpensive routine parameters might be potential components. Currently, liver fat indices lack precision for NAFLD prediction or monitoring in individual patients, but in large cohorts they may substitute nonexistent imaging data and serve as a compound biomarker of metabolic syndrome and its cardiometabolic sequelae.

The Prevalence of Non-Alcoholic Fatty Liver Disease in Iranian Children and Adult Population: A Systematic Review and Meta-Analysis

Background

Non-alcoholic fatty liver disease (NAFLD) is the world's most common etiology of chronic liver disease. In this systematic review and meta-analysis, we estimated the prevalence of NAFLD in the Iranian children and adult population.

Methods

A comprehensive search of five international databases, including PubMed, ISI/WOS, ProQuest, Scopus, and Google Scholar, was done from inception to Nov 2022. Studies on NAFLD patients and their risk factors were selected for meta-analysis. The quality of the included studies was assessed by The Joanna Briggs Institute (JBI) Critical Appraisal Checklist for cross-sectional, and cohort studies. The heterogeneity between studies was investigated using Cochran test and I2 statistics. Random and fixed effect models were used for heterogenic and non-heterogenic studies, respectively. We used Comprehensive Meta-Analysis version 3 for conducting meta-analysis.

Results

Twenty studies were finally included. The total prevalence of NAFLD in children, boys, and girls was 6.7% (95% CI: 0.02-0.18), 12.5% (95% CI: 0.04-0.29) and, 10.1% (95% CI: 0.04-0.21), respectively. The total prevalence of NAFLD in obese children, obese boys, and obese girls was 42% (95% CI: 0.18-0.69), 44% (95% CI: 0.13-0.80), and 33 % (95% CI: 0.13-0.62), respectively. The total prevalence of NAFLD in adults was 36.9% (95% CI: 0.31-0.42). The prevalence of NAFLD in men and women was 33.8% (95% CI: 0.27-0.41) and 29.9% (95% CI: 0.21-0.40), respectively.

Conclusion

NAFLD prevalence in Iranian adults and obese children is considerable; however, data about the children population was insufficient.

Application of machine learning in predicting non-alcoholic fatty liver disease using anthropometric and body composition indices

Non-alcoholic fatty liver disease (NAFLD) is the most common chronic liver disease, which can progress from simple steatosis to advanced cirrhosis and hepatocellular carcinoma. Clinical diagnosis of NAFLD is crucial in the early stages of the disease. The main aim of this study was to apply machine learning (ML) methods to identify significant classifiers of NAFLD using body composition and anthropometric variables. A cross-sectional study was carried out among 513 individuals aged 13 years old or above in Iran. Anthropometric and body composition measurements were performed manually using body composition analyzer InBody 270. Hepatic steatosis and fibrosis were determined using a Fibroscan. ML methods including k-Nearest Neighbor (kNN), Support Vector Machine (SVM), Radial Basis Function (RBF) SVM, Gaussian Process (GP), Random Forest (RF), Neural Network (NN), Adaboost and Naïve Bayes were examined for model performance and to identify anthropometric and body composition predictors of fatty liver disease. RF generated the most accurate model for fatty liver (presence of any stage), steatosis stages and fibrosis stages with 82%, 52% and 57% accuracy, respectively. Abdomen circumference, waist circumference, chest circumference, trunk fat and body mass index were among the most important variables contributing to fatty liver disease. ML-based prediction of NAFLD using anthropometric and body composition data can assist clinicians in decision making. ML-based systems provide opportunities for NAFLD screening and early diagnosis, especially in population-level and remote areas.

Risk Factors of Non-alcoholic Fatty Liver Disease in the Iranian Adult Population: A Systematic Review and Meta-analysis

Context: Non-alcoholic fatty liver disease (NAFLD) is progressing considerably worldwide. Identifying the risk factors of NAFLD is a critical step in preventing its progression. Methods: In November 2022, two independent researchers studied seven databases, including PubMed, ISI/WoS, ProQuest, Scopus, SID, Magiran, and Google Scholar, and reference list of relevant articles, searching studies that assessed NAFLD risk factors in the Iranian adult population. Heterogeneity between studies was assessed by Cochran’s test and its composition using I2 statistics. A random-effects model was used when heterogeneity was observed; otherwise, a fixed-effects model was applied. Egger’s regression test and Trim-and-Fill analysis were used to assess publication bias. Comprehensive Meta-analysis software (version 3) was used for the analyses of the present study. Results: The results of this study showed significant associations between NAFLD with age [n = 15, odds ratio (OR) = 2.12, 95% CI: 1.79 - 2.51], body mass index (n = 46, OR = 5.00, 95% CI: 3.34 - 7.49), waist circumference (n = 20, OR = 6.37, 95% CI: 3.25 - 12.48), waist-to-hip ratio (n = 17, OR = 4.72, 95% CI: 3.93 - 5.66), total cholesterol (n = 39, OR = 1.80, 95% CI: 1.52 - 2.13), high-density lipoprotein (n = 37, OR = 0.53, 95% CI: 0.44 - 0.65), low-density lipoprotein (n = 31, OR = 1.68, 95% CI: 1.38 - 2.05), triglyceride (n = 31, OR = 3.21, 95% CI: 2.67 - 3.87), alanine aminotransferase (n = 26, OR = 4.06, 95% CI: 2.94 - 5.62), aspartate aminotransferase (n = 27, OR = 2.16, 95% CI: 1.50 - 3.12), hypertension (n = 13, OR = 2.53, 95% CI: 2.32 - 2.77), systolic blood pressure (n = 13, OR = 1.83, 95% CI: 1.53 - 2.18), diastolic blood pressure (n = 14, OR = 1.80, 95% CI: 1.48 - 2.20), fasting blood sugar (n = 31,OR = 2.91, 95% CI: 2.11- 4.01), homeostatic model assessment for insulin resistance (n = 5, OR = 1.92, 95% CI: 1.48 - 2.59), diabetes mellitus (n = 15, OR = 3.04, 95% CI: 2.46 - 3.75), metabolic syndrome (n = 10, OR = 3.56, 95% CI: 2.79 - 4.55), and physical activity (n = 11, OR = 0.32, 95% CI: 0.24 - 0.43) (P < 0.05). Conclusions: In conclusion, several factors are significantly associated with NAFLD. However, anthropometric indices had the strongest relationship with NAFLD in the Iranian adult population.

2019 Global NAFLD Prevalence: A Systematic Review and Meta-analysis

BACKGROUND & AIMS

The increasing rates of obesity and type 2 diabetes mellitus may lead to increased prevalence of nonalcoholic fatty liver disease (NAFLD). We aimed to determine the current and recent trends on the global and regional prevalence of NAFLD.

METHODS

Systematic search from inception to March 26, 2020 was performed without language restrictions. Two authors independently performed screening and data extraction. We performed meta-regression to determine trends in NAFLD prevalence.

RESULTS

We identified 17,244 articles from literature search and included 245 eligible studies involving 5,399,254 individuals. The pooled global prevalence of NAFLD was 29.8% (95% confidence interval [CI], 28.6%-31.1%); of these, 82.5% of included articles used ultrasound to diagnose NAFLD, with prevalence of 30.6% (95% CI, 29.2%-32.0%). South America (3 studies, 5716 individuals) and North America (4 studies, 18,236 individuals) had the highest NAFLD prevalence at 35.7% (95% CI, 34.0%-37.5%) and 35.3% (95% CI, 25.4%-45.9%), respectively. From 1991 to 2019, trend analysis showed NAFLD increased from 21.9% to 37.3% (yearly increase of 0.7%, P < .0001), with South America showing the most rapid change of 2.7% per year, followed by Europe at 1.1%.

CONCLUSIONS

Despite regional variation, the global prevalence of NAFLD is increasing overall. Policy makers must work toward reversing the current trends by increasing awareness of NAFLD and promoting healthy lifestyle environments.

Ontology-Based decision tree model for prediction of fatty liver diseases

Non-Alcohol Fatty liver disease is a common clinical complication. The paper aimed to develop a knowledge-based fatty liver detection system based on an ontology and detection rules extracted from a decision tree algorithm. Ontology is created to represent knowledge related to patients and fatty liver disease. By utilizing 43 SWRL rules and the Drool inference engine in ontology, we detected fatty liver patients. The training dataset size is 70% of clean data, including 580 electronic medical records of patients who suffer from liver diseases. After inferencing the rules, the number of patients who suffer from fatty liver disease in ontology is the same as the decision tree model. The paper validated the result generated by the ontology model through the results of the decision tree model.

Sodium in relation with nonalcoholic fatty liver disease: A systematic review and meta-analysis of observational studies

Findings on the association of sodium with nonalcoholic fatty liver disease (NAFLD) are conflicting. The present systematic review and meta‐analysis study aimed to assess the association between salt or sodium intake or serum sodium levels and NAFLD risk. Relevant articles were identified by searching PubMed, Web of Knowledge, Scopus, Proquest, and Embase databases through May 1, 2021, without language restriction. The pooled odds ratio (OR) and 95% confidence interval (CI) were estimated using Der‐Simonian and Laird method and random‐effects meta‐analysis. The certainty of the evidence was rated using the GRADE method. Out of 6470 documents, 7 epidemiological/observational (1 cohort, 1 case–control, and 5 cross‐sectional) studies on the relationship between dietary salt/sodium intakes and NAFLD risk met our inclusion criteria. The meta‐analysis of all studies showed a significant positive association between the highest salt/sodium intake and NALFD risk (OR = 1.60, 95% CI: 1.19–2.15) with a meaningful heterogeneity among studies (I² = 96.70%, p‐value <.001). The NAFLD risk was greater in the studies with higher quality (OR = 1.81, 95% CI: 1.24–2.65) or using the equation‐based methods for NAFLD ascertainment (OR = 2.02, 95% CI: 1.29–3.17) or urinary sodium collection as a sodium intake assessment (OR = 2.48, 95% CI: 1.52–4.06). The overall certainty of the evidence was very low. In conclusion, high sodium intake seems to be related to increased NAFLD risk. Further well‐designed studies are needed to clarify this association and shed light on the underlying mechanisms.

Machine learning models for predicting non-alcoholic fatty liver disease in the general United States population: NHANES database

BACKGROUND

Non-alcoholic fatty liver disease (NAFLD) is the most common chronic liver disease, affecting over 30% of the United States population. Early patient identification using a simple method is highly desirable.

AIM

To create machine learning models for predicting NAFLD in the general United States population.

METHODS

Using the NHANES 1988-1994. Thirty NAFLD-related factors were included. The dataset was divided into the training (70%) and testing (30%) datasets. Twenty-four machine learning algorithms were applied to the training dataset. The best-performing models and another interpretable model (i.e., coarse trees) were tested using the testing dataset.

RESULTS

There were 3235 participants (n = 3235) that met the inclusion criteria. In the training phase, the ensemble of random undersampling (RUS) boosted trees had the highest F1 (0.53). In the testing phase, we compared selective machine learning models and NAFLD indices. Based on F1, the ensemble of RUS boosted trees remained the top performer (accuracy 71.1% and F1 0.56) followed by the fatty liver index (accuracy 68.8% and F1 0.52). A simple model (coarse trees) had an accuracy of 74.9% and an F1 of 0.33.

CONCLUSION

Not every machine learning model is complex. Using a simpler model such as coarse trees, we can create an interpretable model for predicting NAFLD with only two predictors: fasting C-peptide and waist circumference. Although the simpler model does not have the best performance, its simplicity is useful in clinical practice.

Current-Visit and Next-Visit Prediction for Fatty Liver Disease With a Large-Scale Dataset: Model Development and Performance Comparison

BACKGROUND

Fatty liver disease (FLD) arises from the accumulation of fat in the liver and may cause liver inflammation, which, if not well controlled, may develop into liver fibrosis, cirrhosis, or even hepatocellular carcinoma.

OBJECTIVE

We describe the construction of machine-learning models for current-visit prediction (CVP), which can help physicians obtain more information for accurate diagnosis, and next-visit prediction (NVP), which can help physicians provide potential high-risk patients with advice to effectively prevent FLD.

METHODS

The large-scale and high-dimensional dataset used in this study comes from Taipei MJ Health Research Foundation in Taiwan. We used one-pass ranking and sequential forward selection (SFS) for feature selection in FLD prediction. For CVP, we explored multiple models, including k-nearest-neighbor classifier (KNNC), Adaboost, support vector machine (SVM), logistic regression (LR), random forest (RF), Gaussian naïve Bayes (GNB), decision trees C4.5 (C4.5), and classification and regression trees (CART). For NVP, we used long short-term memory (LSTM) and several of its variants as sequence classifiers that use various input sets for prediction. Model performance was evaluated based on two criteria: the accuracy of the test set and the intersection over union/coverage between the features selected by one-pass ranking/SFS and by domain experts. The accuracy, precision, recall, F-measure, and area under the receiver operating characteristic curve were calculated for both CVP and NVP for males and females, respectively.

RESULTS

After data cleaning, the dataset included 34,856 and 31,394 unique visits respectively for males and females for the period 2009-2016. The test accuracy of CVP using KNNC, Adaboost, SVM, LR, RF, GNB, C4.5, and CART was respectively 84.28%, 83.84%, 82.22%, 82.21%, 76.03%, 75.78%, and 75.53%. The test accuracy of NVP using LSTM, bidirectional LSTM (biLSTM), Stack-LSTM, Stack-biLSTM, and Attention-LSTM was respectively 76.54%, 76.66%, 77.23%, 76.84%, and 77.31% for fixed-interval features, and was 79.29%, 79.12%, 79.32%, 79.29%, and 78.36%, respectively, for variable-interval features.

CONCLUSIONS

This study explored a large-scale FLD dataset with high dimensionality. We developed FLD prediction models for CVP and NVP. We also implemented efficient feature selection schemes for current- and next-visit prediction to compare the automatically selected features with expert-selected features. In particular, NVP emerged as more valuable from the viewpoint of preventive medicine. For NVP, we propose use of feature set 2 (with variable intervals), which is more compact and flexible. We have also tested several variants of LSTM in combination with two feature sets to identify the best match for male and female FLD prediction. More specifically, the best model for males was Stack-LSTM using feature set 2 (with 79.32% accuracy), whereas the best model for females was LSTM using feature set 1 (with 81.90% accuracy).

Assessment of Classification Models and Relevant Features on Nonalcoholic Steatohepatitis Using Random Forest

Nonalcoholic fatty liver disease (NAFLD) is the hepatic manifestation of metabolic syndrome and is the most common cause of chronic liver disease in developed countries. Certain conditions, including mild inflammation biomarkers, dyslipidemia, and insulin resistance, can trigger a progression to nonalcoholic steatohepatitis (NASH), a condition characterized by inflammation and liver cell damage. We demonstrate the usefulness of machine learning with a case study to analyze the most important features in random forest (RF) models for predicting patients at risk of developing NASH. We collected data from patients who attended the Cardiovascular Risk Unit of Mostoles University Hospital (Madrid, Spain) from 2005 to 2021. We reviewed electronic health records to assess the presence of NASH, which was used as the outcome. We chose RF as the algorithm to develop six models using different pre-processing strategies. The performance metrics was evaluated to choose an optimized model. Finally, several interpretability techniques, such as feature importance, contribution of each feature to predictions, and partial dependence plots, were used to understand and explain the model to help obtain a better understanding of machine learning-based predictions. In total, 1525 patients met the inclusion criteria. The mean age was 57.3 years, and 507 patients had NASH (prevalence of 33.2%). Filter methods (the chi-square and Mann-Whitney-Wilcoxon tests) did not produce additional insight in terms of interactions, contributions, or relationships among variables and their outcomes. The random forest model correctly classified patients with NASH to an accuracy of 0.87 in the best model and to 0.79 in the worst one. Four features were the most relevant: insulin resistance, ferritin, serum levels of insulin, and triglycerides. The contribution of each feature was assessed via partial dependence plots. Random forest-based modeling demonstrated that machine learning can be used to improve interpretability, produce understanding of the modeled behavior, and demonstrate how far certain features can contribute to predictions.

Liver fat scores do not reflect interventional changes in liver fat content induced by high-protein diets

Non-alcoholic fatty liver disease (NAFLD) is common in Metabolic Syndrome and type 2 diabetes (T2DM), driven by energy imbalance, saturated fats and simple carbohydrates. NAFLD requires screening and monitoring for late complications. Liver fat indices may predict NAFLD avoiding expensive or invasive gold-standard methods, but they are poorly validated for use in interventional settings. Recent data indicate a particular insensitivity to weight-independent liver fat reduction. We evaluated 31 T2DM patients, completing a randomized intervention study on isocaloric high-protein diets. We assessed anthropometric measures, intrahepatic lipid (IHL) content and serum liver enzymes, allowing AUROC calculations as well as cross-sectional and longitudinal Spearman correlations between the fatty liver index, the NAFLD-liver fat score, the Hepatosteatosis Index, and IHL. At baseline, all indices predicted NAFLD with moderate accuracy (AUROC 0.731–0.770), supported by correlation analyses. Diet-induced IHL changes weakly correlated with changes of waist circumference, but no other index component or the indices themselves. Liver fat indices may help to easily detect NAFLD, allowing cost-effective allocation of further diagnostics to patients at high risk. IHL reduction by weight-independent diets is not reflected by a proportional change in liver fat scores. Further research on the development of treatment-sensitive indices is required.

Trial registration: The trial was registered at clinicaltrials.gov: NCT02402985.

Implementation of Combinational Deep Learning Algorithm for Non-alcoholic Fatty Liver Classification in Ultrasound Images

Background

Nowadays, fatty liver is one of the commonly occurred diseases for the liver which can be observed generally in obese patients. Final results from a variety of exams and imaging methods can help to identify and evaluate people affected by this condition.

Objective

The aim of this study is to present a combined algorithm based on neural networks for the classification of ultrasound ‎images from fatty liver affected patients.

Material and Methods

In experimental research can be categorized as a diagnostic study which focuses on classification of the acquired ultrasonography images for 55 patients with fatty liver. We implemented pre-trained convolutional neural networks of Inception-ResNetv2, GoogleNet, AlexNet, and ResNet101 to extract features from the images and after combining these resulted features, we provided support vector machine (SVM) algorithm to classify the liver images. Then the results are compared with the ones in implementing the algorithms independently.

Results

The area under the receiver operating characteristic curve (AUC) for the introduced combined network resulted in 0.9999, which is a better result compared to any of the other introduced algorithms. The resulted accuracy for the proposed network also caused 0.9864, which seems acceptable accuracy for clinical application.

Conclusion

The proposed network can be used with high accuracy to classify ultrasound images of the liver to normal or fatty. The presented approach besides the high AUC in comparison with other methods have the independence of the method from the ‎user or expert interference.

Non-alcoholic Fatty Liver and Liver Fibrosis Predictive Analytics: Risk Prediction and Machine Learning Techniques for Improved Preventive Medicine

Non-alcoholic fatty liver disease (NAFLD) is the most common liver disease worldwide, with a prevalence of 20%-30% in the general population. NAFLD is associated with increased risk of cardiovascular disease and may progress to cirrhosis with time. The purpose of this study was to predict the risks associated with NAFLD and advanced fibrosis on the Fatty Liver Index (FLI) and the 'NAFLD fibrosis 4' calculator (FIB-4), to enable physicians to make more optimal preventive medical decisions. A prospective cohort of apparently healthy volunteers from the Tel Aviv Medical Center Inflammation Survey (TAMCIS), admitted for their routine annual health check-up. Data from the TAMCIS database were subjected to machine learning classification models to predict individual risk after extensive data preparation that included the computation of independent variables over several time points. After incorporating the time covariates and other key variables, this technique outperformed the predictive power of current popular methods (an improvement in AUC above 0.82). New powerful factors were identified during the predictive process. The findings can be used for risk stratification and in planning future preventive strategies based on lifestyle modifications and medical treatment to reduce the disease burden. Interventions to prevent chronic disease can substantially reduce medical complications and the costs of the disease. The findings highlight the value of predictive analytic tools in health care environments. NAFLD constitutes a growing burden on the health system; thus, identification of the factors related to its incidence can make a strong contribution to preventive medicine.

Liver Resection for Hepatocellular Carcinoma in Non-alcoholic Fatty Liver Disease: a Multicenter Propensity Matching Analysis with HBV-HCC

BACKGROUND

The incidence of hepatocellular carcinoma (HCC) in non-alcoholic fatty liver disease (NAFLD) is increasing worldwide. Higher perioperative risks may be anticipated due to underlying steatohepatitis, while long-term outcomes after liver resection are unknown. We sought to investigate outcomes after liver resection for NAFLD-HCC versus hepatitis B virus (HBV)-HCC using propensity score matching (PSM).

METHODS

Consecutive patients who underwent liver resection for HCC between 2003 and 2014 were identified from a multicenter database. Patients with NAFLD-HCC were matched one-to-one to patients with HBV-HCC.

RESULTS

Among 1483 patients identified, 96 (6.5%) had NAFLD-HCC and 1387 (93.5%) had HBV-HCC. Patients with NAFLD-HCC were older (median age 57 vs. 50 years), more often overweight (50.0% vs. 37.5%), less often to have cirrhosis (30.2% vs. 72.5%) and liver dysfunction (Child-Pugh B: 4.2% vs. 10.7%), had larger tumor size (median 7.2 vs. 6.2 cm) yet had better tumor differentiation (27.1% vs. 17.6%) compared with patients with HBV-HCC (all P < 0.05). Perioperative mortality and morbidity were comparable between the two groups (1.0% vs. 1.4% and 20.8% vs. 23.2%, both P > 0.05). No differences were noted in median OS and RFS among patient with NAFLD-HCC versus HBV-HCC before or after PSM.

CONCLUSION

While patients with NAFLD-HCC had different clinical characteristics than patients with HBV-HCC, liver resection resulted in similar perioperative outcomes and comparable OS and RFS among patients with NAFLD-HCC and HBV-HCC.

[Predictive value of body mass index combined with waist circumference for new-onset nonalcoholic fatty liver disease in patients with type 2 diabetes mellitus]

OBJECTIVE

To investigate the predictive value of body mass index (BMI) combined with waist circumference (WC) for new-onset nonalcoholic fatty liver disease (NAFLD) in patients with type 2 diabetes mellitus (T2DM).

METHODS

This community-based prospective cohort study was conducted among 3501 T2DM patients without NAFLD recruited from the staff of Kailuan Company, who underwent routine physical examination in the year 2006 and 2007, and a total of 2920 subjects were included in the final analysis. According to the baseline BMI and WC, the subjects were divided into group A (with normal BMI and WC), group B (with normal BMI but elevated WC), group C (with elevated BMI but a normal WC) and group D (with elevated BMI and WC). The subjects in the 4 groups were followed for the occurrence of NAFLD by reviewing their reports of physical examinations during the periods of 2008-2009, 2010-2011, 2012-2013, 2014-2015 and 2016-2017. The cumulative incidence of NAFLD was compared across the 4 groups and Cox regression analysis was used to test the correlation of BMI and WC with new onset of NAFLD.

RESULTS

The cumulative incidence of NAFLD increased progressively in the 4 groups (50%, 66%, 68% and 77%, respectively). Cox regression analysis showed that compared with group A, groups B, C and D had increased risks of NAFLD after adjusting for age, gender and other risk factors, with HR values of 1.62, 1.98 and 2.47, respectively.

CONCLUSIONS

Elevated BMI and WC are both independent risk factors for NAFLD in type 2 diabetic patients, and the combination of BMI and WC has a greater predictive value for NAFLD than either of them alone.

Prediction of fatty liver disease using machine learning algorithms

BACKGROUND AND OBJECTIVE

Fatty liver disease (FLD) is a common clinical complication; it is associated with high morbidity and mortality. However, an early prediction of FLD patients provides an opportunity to make an appropriate strategy for prevention, early diagnosis and treatment. We aimed to develop a machine learning model to predict FLD that could assist physicians in classifying high-risk patients and make a novel diagnosis, prevent and manage FLD.

METHODS

We included all patients who had an initial fatty liver screening at the New Taipei City Hospital between 1st and 31st December 2009. Classification models such as random forest (RF), Naïve Bayes (NB), artificial neural networks (ANN), and logistic regression (LR) were developed to predict FLD. The area under the receiver operating characteristic curve (ROC) was used to evaluate performances among the four models.

RESULTS

A total of 577 patients were included in this study; of those 377 patients had fatty liver. The area under the receiver operating characteristic (AUROC) of RF, NB, ANN, and LR with 10 fold-cross validation was 0.925, 0.888, 0.895, and 0.854 respectively. Additionally, The accuracy of RF, NB, ANN, and LR 87.48, 82.65, 81.85, and 76.96%.

CONCLUSION

In this study, we developed and compared the four classification models to predict fatty liver disease accurately. However, the random forest model showed higher performance than other classification models. Implementation of a random forest model in the clinical setting could help physicians to stratify fatty liver patients for primary prevention, surveillance, early treatment, and management.

Liver Fat Scores Moderately Reflect Interventional Changes in Liver Fat Content by a Low-Fat Diet but Not by a Low-Carb Diet

Background: Non-alcoholic fatty liver disease (NAFLD) is a common metabolic disorder all over the world, mainly being associated with a sedentary lifestyle, adiposity, and nutrient imbalance. The increasing prevalence of NAFLD accommodates similar developments for type 2 diabetes and diabetes-related comorbidities and complications. Therefore, early detection of NAFLD is an utmost necessity. Potentially helpful tools for the prediction of NAFLD are liver fat indices. The fatty liver index (FLI) and the NAFLD-liver fat score (NAFLD-LFS) have been recently introduced for this aim. However, both indices have been shown to correlate with liver fat status, but there is neither sufficient data on the longitudinal representation of liver fat change, nor proof of a diet-independent correlation between actual liver fat change and change of index values. While few data sets on low-fat diets have been published recently, low-carb diets have not been yet assessed in this context. Aim: We aim to provide such data from a highly effective short-term intervention to reduce liver fat, comparing a low-fat and a low-carb diet in subjects with prediabetes. Methods: Anthropometric measurements, magnetic resonance (MR)-based intrahepatic lipid (IHL) content, and several serum markers for liver damage have been collected in 140 subjects, completing the diet phase in this trial. Area-under-the-responder-operator-curves (AUROC) calculations as well as cross-sectional and longitudinal Spearman correlations were used. Results: Both FLI and NAFLD-LFS predict liver fat with moderate accuracy at baseline (AUROC 0.775–0.786). These results are supported by correlation analyses. Changes in liver fat, achieved by the dietary intervention, correlate moderately with changes in FLI and NAFLD-LFS in the low-fat diet, but not in the low-carb diet. A correlation analysis between change of actual IHL content and change of single elements of the liver fat indices revealed diet-specific moderate to strong correlations between ΔIHL and changes of measures of obesity, ΔTG, and ΔALT (all low-fat, only) and between ΔIHL and ΔGGT (low-carb, only). With exception for a stronger decrease of triglycerides (TG) levels in the low-carb diet, there is no statistically significant difference in the effect of the diets on anthropometric or serum-based score parameters. Conclusion: While liver fat indices have proved useful in the early detection of NAFLD and may serve as a cost-saving substitute for expensive MR measurements in the cross-sectional evaluation of liver status, their capability to represent interventional changes of liver fat content appears to be diet-specific and lacks accuracy. Liver fat reduction by low-fat diets can be monitored with moderate precision, while low-carb diets require different measuring techniques to demonstrate the same dietary effect.