Non-invasive diagnosis of non-alcoholic steatohepatitis and fibrosis with the use of omics and supervised learning: A proof of concept study

Exploring Biomarkers in Nonalcoholic Fatty Liver Disease Among Individuals With Type 2 Diabetes Mellitus

Integrating biomarkers into a comprehensive strategy is crucial for precise patient management, especially considering the significant healthcare costs associated with diseases. Current studies emphasize the urgent need for a paradigm shift in conceptualizing nonalcoholic fatty liver disease (NAFLD), now renamed metabolic dysfunction-associated steatotic liver disease (MASLD). Biomarkers are emerging as indispensable tools for accurate diagnosis, risk stratification, and monitoring disease progression. This review classifies biomarkers into conventional and novel categories, such as lipids, insulin resistance, hepatic function, and cutting-edge imaging/omics, and evaluates their potential to transform the approach to MASLD among individuals with type 2 diabetes mellitus (T2D). It focuses on the critical role of biomarkers in early MASLD detection, enhancing predictive accuracy, and discerning responses to interventions (pharmacological or lifestyle modifications). Amid this discussion, the complexities of the relationship between T2D and MASLD are explored, considering factors like age, gender, genetics, ethnicity, and socioeconomic background. Biomarkers enhance the effectiveness of interventions and support global initiatives to reduce the burden of MASLD, thereby improving public health outcomes. This review recognizes the promising potential of biomarkers for diagnostic precision while candidly addressing the challenges in implementing these advancements in clinical practice. The transformative role of biomarkers emerges as a central theme, promising to reshape our understanding of disease trajectories, prognosis, and the customization of personalized therapeutic strategies for improved patient outcomes. From a future perspective, identifying early-stage biomarkers, understanding environmental impact through exposomes, and applying a multiomics approach may reveal additional insight into MASLD development.

Lipid metabolism in MASLD and MASH: From mechanism to the clinic

Metabolic dysfunction-associated steatotic liver disease/steatohepatitis (MASLD/MASH) is recognised as a metabolic disease characterised by excess intrahepatic lipid accumulation due to lipid overflow and synthesis, alongside impaired oxidation and/or export of these lipids. But where do these lipids come from? The main pathways related to hepatic lipid accumulation are de novo lipogenesis and excess fatty acid transport to the liver (due to increased lipolysis, adipose tissue insulin resistance, as well as excess dietary fatty acid intake, in particular of saturated fatty acids). Not only triglycerides but also other lipids are secreted by the liver and are associated with a worse histological profile in MASH, as shown by lipidomics. Herein, we review the role of lipid metabolism in MASLD/MASH and discuss the impact of weight loss (diet, bariatric surgery, GLP-1RAs) or other pharmacological treatments (PPAR or THRβ agonists) on hepatic lipid metabolism, lipidomics, and the resolution of MASH.

A specific serum lipid signature characterizes patients with glycogen storage disease type Ia

Glycogen storage disease type Ia (GSDIa) is a rare, inherited glucose-6-phosphatase-α (G6Pase-α) deficiency-induced carbohydrate metabolism disorder. Although hyperlipidemia is a hallmark of GSDI, the extent of lipid metabolism disruption remains incompletely understood. Lipidomic analysis was performed to characterize the serum lipidome in patients with GSDIa, by including age- and sex-matched healthy controls and age-matched hypercholesterolemic controls. Metabolic control and dietary information biochemical markers were obtained from patients with GSDIa. Patients with GSDIa showed higher total serum lysophosphatidylcholine (Fold Change, (FC) 2.2, P < 0.0001), acyl-acyl-phosphatidylcholine (FC 2.1, P < 0.0001), and ceramide (FC 2.4, P < 0.0001) levels and bile acid (FC 0.7, P < 0.001), acylcarnitines (FC 0.7, P < 0.001), and cholesterol esters (FC 1.0, P < 0.001) than those of healthy controls, and higher di- (FC 1.1, P < 0.0001; FC 0.9, P < 0.01) and triacylglycerol (FC 6.3, P < 0.0001; FC 3.9, P < 0.01) levels than those of healthy controls and hypercholesterolemic subjects. Both total cholesterol and triglyceride values correlated with Cer (d16:1/22:0), Cer (d18:1/20:0), Cer (d18:1/20:0(OH)), Cer (d18:1/22:0), Cer (d18:1/23:0), Cer (d18:1/24:1), Cer (d18:2/22:0), Cer (d18:2/24:1). Total cholesterol also correlated with Cer (d18:1/24:0), Cer (d18:2/20:0), HexCer (d16:1/22:0), HexCer (d18:1/18:0), and Hex2Cer (d18:1/20:0). Triglyceridelevels correlated with Cer (d18:0/24:1). Alanine aminotransferase values correlated with Cer (d18:0/22:0), insulin with Cer (d18:1/22:1) and Cer (d18:1/24:1), and HDL with hexosylceramide (HexCer) (d18:2/23:0). These results expand on the currently known involvement of lipid metabolism in GSDIa. Circulating Cer may allow for refined dietary assessment compared with traditional biomarkers. Because specific lipid species are relatively easy to assess, they represent potential novel biomarkers of GSDIa.

Development and validation of a machine learning-based framework for assessing metabolic-associated fatty liver disease risk

BACKGROUND

The existing predictive models for metabolic-associated fatty liver disease (MAFLD) possess certain limitations that render them unsuitable for extensive population-wide screening. This study is founded upon population health examination data and employs a comparison of eight distinct machine learning (ML) algorithms to construct the optimal screening model for identifying high-risk individuals with MAFLD in China.

METHODS

We collected physical examination data from 5,171,392 adults residing in the northwestern region of China, during the year 2021. Feature selection was conducted through the utilization of the Least Absolute Shrinkage and Selection Operator (LASSO) regression. Additionally, class balancing parameters were incorporated into the models, accompanied by hyperparameter tuning, to effectively address the challenges posed by imbalanced datasets. This study encompassed the development of both tree-based ML models (including Classification and Regression Trees, Random Forest, Adaptive Boosting, Light Gradient Boosting Machine, Extreme Gradient Boosting, and Categorical Boosting) and alternative ML models (specifically, k-Nearest Neighbors and Artificial Neural Network) for the purpose of identifying individuals with MAFLD. Furthermore, we visualized the importance scores of each feature on the selected model.

RESULTS

The average age (standard deviation) of the 5,171,392 participants was 51.12 (15.00) years, with 52.47% of the participants being females. MAFLD was diagnosed by specialized physicians. 20 variables were finally included for analyses after LASSO regression model. Following ten rounds of cross-validation and parameter optimization for each algorithm, the CatBoost algorithm exhibited the best performance, achieving an Area Under the Receiver Operating Characteristic Curve (AUC) of 0.862. The ranking of feature importance indicates that age, BMI, triglyceride, fasting plasma glucose, waist circumference, occupation, high density lipoprotein cholesterol, low density lipoprotein cholesterol, total cholesterol, systolic blood pressure, diastolic blood pressure, ethnicity and cardiovascular diseases are the top 13 crucial factors for MAFLD screening.

CONCLUSION

This study utilized a large-scale, multi-ethnic physical examination data from the northwestern region of China to establish a more accurate and effective MAFLD risk screening model, offering a new perspective for the prediction and prevention of MAFLD.

Artificial intelligence applied to 'omics data in liver disease: towards a personalised approach for diagnosis, prognosis and treatment

Advancements in omics technologies and artificial intelligence (AI) methodologies are fuelling our progress towards personalised diagnosis, prognosis and treatment strategies in hepatology. This review provides a comprehensive overview of the current landscape of AI methods used for analysis of omics data in liver diseases. We present an overview of the prevalence of different omics levels across various liver diseases, as well as categorise the AI methodology used across the studies. Specifically, we highlight the predominance of transcriptomic and genomic profiling and the relatively sparse exploration of other levels such as the proteome and methylome, which represent untapped potential for novel insights. Publicly available database initiatives such as The Cancer Genome Atlas and The International Cancer Genome Consortium have paved the way for advancements in the diagnosis and treatment of hepatocellular carcinoma. However, the same availability of large omics datasets remains limited for other liver diseases. Furthermore, the application of sophisticated AI methods to handle the complexities of multiomics datasets requires substantial data to train and validate the models and faces challenges in achieving bias-free results with clinical utility. Strategies to address the paucity of data and capitalise on opportunities are discussed. Given the substantial global burden of chronic liver diseases, it is imperative that multicentre collaborations be established to generate large-scale omics data for early disease recognition and intervention. Exploring advanced AI methods is also necessary to maximise the potential of these datasets and improve early detection and personalised treatment strategies.

Effects of Combined Low-Dose Spironolactone Plus Vitamin E versus Vitamin E Monotherapy on Lipidomic Profile in Non-Alcoholic Fatty Liver Disease: A Post Hoc Analysis of a Randomized Controlled Trial

Background/Objectives: Lipid dysmetabolism seems to contribute to the development and progression of nonalcoholic fatty liver disease (NAFLD). Our aim was to compare serum lipidomic profile between patients with NAFLD having received monotherapy with vitamin E (400 IU/d) and those having received combination therapy with vitamin E (400 IU/d) and low-dose spironolactone (25 mg/d) for 52 weeks. Methods: This was a post hoc study of a randomized controlled trial (NCT01147523). Serum lipidomic analysis was performed in vitamin E monotherapy group (n = 15) and spironolactone plus vitamin E combination therapy group (n = 12). We employed an untargeted liquid chromatography-mass spectrometry lipid profiling approach in positive and negative ionization mode. Results: Univariate analysis revealed 36 lipid molecules statistically different between groups in positive mode and seven molecules in negative mode. Multivariate analysis in negative mode identified six lipid molecules that remained robustly different between groups. After adjustment for potential confounders, including gender, omega-3 supplementation, leptin concentration and homeostasis model assessment-insulin resistance (HOMA-IR), four lipid molecules remained significant between groups: FA 20:5, SM 34:2;O2, SM 42:3;O2 and CE 22:6, all being higher in the combination treatment group. Conclusions: The combination of spironolactone with vitamin E led to higher circulating levels of four lipid molecules than vitamin E monotherapy, after adjustment for potential confounders. Owing to very limited relevant data, we could not support that these changes in lipid molecules may be beneficial or not for the progression of NAFLD. Thus, mechanistic studies are warranted to clarify the potential clinical significance of these findings.

Predicting Non-Alcoholic Steatohepatitis: A Lipidomics-Driven Machine Learning Approach

Nonalcoholic fatty liver disease (NAFLD), nowadays the most prevalent chronic liver disease in Western countries, is characterized by a variable phenotype ranging from steatosis to nonalcoholic steatohepatitis (NASH). Intracellular lipid accumulation is considered the hallmark of NAFLD and is associated with lipotoxicity and inflammation, as well as increased oxidative stress levels. In this study, a lipidomic approach was used to investigate the plasma lipidome of 12 NASH patients, 10 Nonalcoholic Fatty Liver (NAFL) patients, and 15 healthy controls, revealing significant alterations in lipid classes, such as glycerolipids and glycerophospholipids, as well as fatty acid compositions in the context of steatosis and steatohepatitis. A machine learning XGBoost algorithm identified a panel of 15 plasma biomarkers, including HOMA-IR, BMI, platelets count, LDL-c, ferritin, AST, FA 12:0, FA 18:3 ω3, FA 20:4 ω6/FA 20:5 ω3, CAR 4:0, LPC 20:4, LPC O-16:1, LPE 18:0, DG 18:1_18:2, and CE 20:4 for predicting steatohepatitis. This research offers insights into the connection between imbalanced lipid metabolism and the formation and progression of NAFL D, while also supporting previous research findings. Future studies on lipid metabolism could lead to new therapeutic approaches and enhanced risk assessment methods, as the shift from isolated steatosis to NASH is currently poorly understood.

Circulating total and H-specific GDF15 levels are elevated in subjects with MASLD but not in hyperlipidemic but otherwise metabolically healthy subjects with obesity

BACKGROUND

Growth differentiation factor 15 (GDF15) is a mitokine, the role of which, total or H-specific, in modulating energy metabolism and homeostasis in obesity-related diseases, such as metabolic dysfunction associated steatotic liver disease (MASLD), has not been fully elucidated in adult humans. We aimed to investigate the fasting and stimulated levels of GDF15, total and H-specific, glucose-dependent insulinotropic polypeptide (GIP) and C-peptide, in two physiology interventional studies: one focusing on obesity, and the other on MASLD.

METHODS

Study 1 investigated individuals with normal weight or with obesity, undergoing a 3-h mixed meal test (MMT); and study 2, examined adults with MASLD and controls undergoing a 120-min oral glucose tolerance test (OGTT). Exploratory correlations of total and H-specific GDF15 with clinical, hormonal and metabolomic/lipidomic parameters were also performed.

RESULTS

In study 1, 15 individuals were included per weight group. Fasting and postprandial total and H-specific GDF15 were similar between groups, whereas GIP was markedly higher in leaner individuals and was upregulated following a MMT. Baseline and postprandial C-peptide were markedly elevated in people with obesity compared with lean subjects. GIP was higher in leaner individuals and was upregulated after a MMT, while C-peptide and its overall AUC after a MMT was markedly elevated in people with obesity compared with lean subjects. In study 2, 27 individuals were evaluated. Fasting total GDF15 was similar, but postprandial total GDF15 levels were significantly higher in MASLD patients compared to controls. GIP and C-peptide remained unaffected. The postprandial course of GDF15 was clustered among those of triglycerides and molecules of the alanine cycle, was robustly elevated under MASLD, and constituted the most notable differentiating molecule between healthy and MASLD status. We also present robust positive correlations of the incremental area under the curve of total and H-specific GDF15 with a plethora of lipid subspecies, which remained significant after adjusting for confounders.

CONCLUSION

Serum GDF15 levels do not differ in relation to weight status in hyperlipidemic but otherwise metabolically healthy individuals. In contrast, GDF15 levels are significantly increased in MASLD patients at baseline and they remain significantly higher compared to healthy participants during OGTT, pointing to a role for GDF15 as a mitokine with important roles in the pathophysiology and possibly therapeutics of MASLD. Trial registration ClinicalTrials.gov NCT03986684, NCT04430946.

Steatotic Liver Disease: Pathophysiology and Emerging Pharmacotherapies

Steatotic liver disease (SLD) displays a dynamic and complex disease phenotype. Consequently, the metabolic dysfunction-associated steatotic liver disease (MASLD)/metabolic dysfunction-associated steatohepatitis (MASH) therapeutic pipeline is expanding rapidly and in multiple directions. In parallel, noninvasive tools for diagnosing and monitoring responses to therapeutic interventions are being studied, and clinically feasible findings are being explored as primary outcomes in interventional trials. The realization that distinct subgroups exist under the umbrella of SLD should guide more precise and personalized treatment recommendations and facilitate advancements in pharmacotherapeutics. This review summarizes recent updates of pathophysiology-based nomenclature and outlines both effective pharmacotherapeutics and those in the pipeline for MASLD/MASH, detailing their mode of action and the current status of phase 2 and 3 clinical trials. Of the extensive arsenal of pharmacotherapeutics in the MASLD/MASH pipeline, several have been rejected, whereas other, mainly monotherapy options, have shown only marginal benefits and are now being tested as part of combination therapies, yet others are still in development as monotherapies. Although the Food and Drug Administration (FDA) has recently approved resmetirom, additional therapeutic approaches in development will ideally target MASH and fibrosis while improving cardiometabolic risk factors. Due to the urgent need for the development of novel therapeutic strategies and the potential availability of safety and tolerability data, repurposing existing and approved drugs is an appealing option. Finally, it is essential to highlight that SLD and, by extension, MASLD should be recognized and approached as a systemic disease affecting multiple organs, with the vigorous implementation of interdisciplinary and coordinated action plans. SIGNIFICANCE STATEMENT: Steatotic liver disease (SLD), including metabolic dysfunction-associated steatotic liver disease and metabolic dysfunction-associated steatohepatitis, is the most prevalent chronic liver condition, affecting more than one-fourth of the global population. This review aims to provide the most recent information regarding SLD pathophysiology, diagnosis, and management according to the latest advancements in the guidelines and clinical trials. Collectively, it is hoped that the information provided furthers the understanding of the current state of SLD with direct clinical implications and stimulates research initiatives.

Spatial metabolomics and its application in the liver

Hepatocytes work in highly structured, repetitive hepatic lobules. Blood flow across the radial axis of the lobule generates oxygen, nutrient, and hormone gradients, which result in zoned spatial variability and functional diversity. This large heterogeneity suggests that hepatocytes in different lobule zones may have distinct gene expression profiles, metabolic features, regenerative capacity, and susceptibility to damage. Here, we describe the principles of liver zonation, introduce metabolomic approaches to study the spatial heterogeneity of the liver, and highlight the possibility of exploring the spatial metabolic profile, leading to a deeper understanding of the tissue metabolic organization. Spatial metabolomics can also reveal intercellular heterogeneity and its contribution to liver disease. These approaches facilitate the global characterization of liver metabolic function with high spatial resolution along physiological and pathological time scales. This review summarizes the state of the art for spatially resolved metabolomic analysis and the challenges that hinder the achievement of metabolome coverage at the single-cell level. We also discuss several major contributions to the understanding of liver spatial metabolism and conclude with our opinion on the future developments and applications of these exciting new technologies.

Artificial intelligence in endocrinology: a comprehensive review

BACKGROUND AND AIM

Artificial intelligence (AI) has emerged as a promising technology in the field of endocrinology, offering significant potential to revolutionize the diagnosis, treatment, and management of endocrine disorders. This comprehensive review aims to provide a concise overview of the current landscape of AI applications in endocrinology and metabolism, focusing on the fundamental concepts of AI, including machine learning algorithms and deep learning models.

METHODS

The review explores various areas of endocrinology where AI has demonstrated its value, encompassing screening and diagnosis, risk prediction, translational research, and "pre-emptive medicine". Within each domain, relevant studies are discussed, offering insights into the methodology and main findings of AI in the treatment of different pathologies, such as diabetes mellitus and related disorders, thyroid disorders, adrenal tumors, and bone and mineral disorders.

RESULTS

Collectively, these studies show the valuable contributions of AI in optimizing healthcare outcomes and unveiling new understandings of the intricate mechanisms underlying endocrine disorders. Furthermore, AI-driven approaches facilitate the development of precision medicine strategies, enabling tailored interventions for patients based on their individual characteristics and needs.

CONCLUSIONS

By embracing AI in endocrinology, a future can be envisioned where medical professionals and AI systems synergistically collaborate, ultimately enhancing the lives of individuals affected by endocrine disorders.

Nonalcoholic fatty liver disease test: an external validation cohort

PURPOSE

Non-invasive diagnosis of nonalcoholic fatty liver disease (NAFLD) and its advanced phenotypes (e.g., nonalcoholic steatohepatitis; NASH) is a hot research topic. The aim of this report was the validation of a novel non-invasive index of NAFLD, the "NAFLD test," recently introduced for the diagnosis of NAFLD (vs. non-NAFLD controls).

METHODS

This was a post-hoc analysis of a previous study. The NAFLD test was calculated in NAFLD patients and non-NAFLD controls; the performance of the test was compared with that of other non-invasive indices of NAFLD (fatty liver index [FLI] and hepatic steatosis index [HSI]), and other indices of NASH (index of NASH [ION] and cytokeratin-18/homeostasis model assessment-insulin resistance/aspartate transaminase index [CHAI]).

RESULTS

The NAFLD test was higher in NAFLD patients than in controls (1.89 ± 0.14 vs. 1.30 ± 0.06, respectively; p < 0.001). In NAFLD patients, the NAFLD test was higher in NASH patients than in those with simple nonalcoholic fatty liver (NAFL) (2.21 ± 0.24 vs. 1.57 ± 0.08, respectively; p = 0.007). The area under the receiver operating characteristic curve (AUC) of the NAFLD test was 0.84 (95% CI: 0.74-0.94; p < 0.001) for differentiation between NAFLD and non-NAFLD controls and its performance was similar to that for FLI and HSI. For differentiation between NASH and NAFL patients, the AUC of the NAFLD test was 0.88 (95% CI: 0.62-0.96; p = 0.007) and its performance was superior to that for ION and CHAI.

CONCLUSIONS

The NAFLD test was validated in this external cohort for the non-invasive diagnosis of NAFLD patients vs. non-NAFLD individuals. It was also shown to differentiate between NASH and NAFL patients with acceptable accuracy.

Artificial Intelligence in Liver Diseases: Recent Advances

Liver diseases cause a significant burden on public health worldwide. In spite of great advances during recent years, there are still many challenges in the diagnosis and treatment of liver diseases. During recent years, artificial intelligence (AI) has been widely used for the diagnosis, risk stratification, and prognostic prediction of various diseases based on clinical datasets and medical images. Accumulative studies have shown its performance for diagnosing patients with nonalcoholic fatty liver disease and liver fibrosis and assessing their severity, and for predicting treatment response and recurrence of hepatocellular carcinoma, outcomes of liver transplantation recipients, and risk of drug-induced liver injury. Herein, we aim to comprehensively summarize the current evidence regarding diagnostic, prognostic, and/or therapeutic role of AI in these common liver diseases.

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.

Predicting prognosis outcomes of primary central nervous system lymphoma with high-dose methotrexate-based chemotherapeutic treatment using lipidomics

Background

Primary central nervous system lymphoma (PCNSL) is a rare extranodal lymphomatous malignancy which is commonly treated with high-dose methotrexate (HD-MTX)-based chemotherapy. However, the prognosis outcome of HD-MTX-based treatment cannot be accurately predicted using the current prognostic scoring systems, such as the Memorial Sloan-Kettering Cancer Center (MSKCC) score.

Methods

We studied 2 cohorts of patients with PCNSL and applied lipidomic analysis to their cerebrospinal fluid (CSF) samples. After removing the batch effects and features engineering, we applied and compared several classic machine-learning models based on lipidomic data of CSF to predict the relapse of PCNSL in patients who were treated with HD-MTX-based chemotherapy.

Results

We managed to remove the batch effects and get the optimum features of each model. Finally, we found that Cox regression had the best prediction performance (AUC = 0.711) on prognosis outcomes.

Conclusions

We developed a Cox regression model based on lipidomic data, which could effectively predict PCNSL patient prognosis before the HD-MTX-based chemotherapy treatments.

Development and validation of machine learning models for nonalcoholic fatty liver disease

BACKGROUND

Nonalcoholic fatty liver disease (NAFLD) had become the most prevalent liver disease worldwide. Early diagnosis could effectively reduce NAFLD-related morbidity and mortality. This study aimed to combine the risk factors to develop and validate a novel model for predicting NAFLD.

METHODS

We enrolled 578 participants completing abdominal ultrasound into the training set. The least absolute shrinkage and selection operator (LASSO) regression combined with random forest (RF) was conducted to screen significant predictors for NAFLD risk. Five machine learning models including logistic regression (LR), RF, extreme gradient boosting (XGBoost), gradient boosting machine (GBM), and support vector machine (SVM) were developed. To further improve model performance, we conducted hyperparameter tuning with train function in Python package 'sklearn'. We included 131 participants completing magnetic resonance imaging into the testing set for external validation.

RESULTS

There were 329 participants with NAFLD and 249 without in the training set, while 96 with NAFLD and 35 without were in the testing set. Visceral adiposity index, abdominal circumference, body mass index, alanine aminotransferase (ALT), ALT/AST (aspartate aminotransferase), age, high-density lipoprotein cholesterol (HDL-C) and elevated triglyceride (TG) were important predictors for NAFLD risk. The area under curve (AUC) of LR, RF, XGBoost, GBM, SVM were 0.915 [95% confidence interval (CI): 0.886-0.937], 0.907 (95% CI: 0.856-0.938), 0.928 (95% CI: 0.873-0.944), 0.924 (95% CI: 0.875-0.939), and 0.900 (95% CI: 0.883-0.913), respectively. XGBoost model presented the best predictive performance, and its AUC was enhanced to 0.938 (95% CI: 0.870-0.950) with further parameter tuning.

CONCLUSIONS

This study developed and validated five novel machine learning models for NAFLD prediction, among which XGBoost presented the best performance and was considered a reliable reference for early identification of high-risk patients with NAFLD in clinical practice.

Recent advances in LC-MS-based metabolomics for clinical biomarker discovery

The employment of liquid chromatography-mass spectrometry (LC-MS) untargeted and targeted metabolomics has led to the discovery of novel biomarkers and improved the understanding of various disease mechanisms. Numerous strategies have been reported to expand the metabolite coverage in LC-MS-untargeted and targeted metabolomics. To improve the sensitivity of low-abundance or poor-ionized metabolites for reducing the amount of clinical sample, chemical derivatization methods are used to target different functional groups. Proper sample preparation is beneficial for reducing the matrix effect, maintaining the stability of the LC-MS system, and increasing the metabolite coverage. Machine learning has recently been integrated into the workflow of LC-MS metabolomics to accelerate metabolite identification and data-processing automation, and increase the accuracy of disease classification and clinical outcome prediction. Due to the rapidly growing utility of LC-MS metabolomics in discovering disease markers, this review will address the recent advances in the field and offer perspectives on various strategies for expanding metabolite coverage, chemical derivatization, sample preparation, clinical disease markers, and machining learning for disease modeling.

Liver fat as risk factor of hepatic and cardiometabolic diseases

Non-alcoholic fatty liver disease (NAFLD) is a disorder characterized by excessive accumulation of fat in the liver that can progress to liver inflammation (non-alcoholic steatohepatitis [NASH]), liver fibrosis, and cirrhosis. Although most efforts for drug development are focusing on the treatment of the latest stages of NAFLD, where significant fibrosis and NASH are present, findings from studies suggest that the amount of liver fat may be an important independent risk factor and/or predictor of development and progression of NAFLD and metabolic diseases. In this review, we first describe the current tools available for quantification of liver fat in humans and then present the clinical and pathophysiological evidence that link liver fat with NAFLD progression as well as with cardiometabolic diseases. Finally, we discuss current pharmacological and non-pharmacological approaches to reduce liver fat and present open questions that have to be addressed in future studies.

Predictive value of radiomics analysis of enhanced CT for three-tiered microvascular invasion grading in hepatocellular carcinoma

BACKGROUND

Microvascular invasion (MVI) is a major risk factor, for recurrence and metastasis of hepatocellular carcinoma (HCC) after radical surgery and liver transplantation. However, its diagnosis depends on the pathological examination of the resected specimen after surgery; therefore, predicting MVI before surgery is necessary to provide reference value for clinical treatment. Meanwhile, predicting only the existence of MVI is not enough, as it ignores the degree, quantity, and distribution of MVI and may lead to MVI-positive patients suffering due to inappropriate treatment. Although some studies have involved M2 (high risk of MVI), majority have adopted the binary classification method or have not included radiomics.

PURPOSE

To develop three-class classification models for predicting the grade of MVI of HCC by combining enhanced computed tomography radiomics features with clinical risk factors.

METHODS

The data of 166 patients with HCC confirmed by surgery and pathology were analyzed retrospectively. The patients were divided into the training (116 cases) and test (50 cases) groups at a ratio of 7:3. Of them, 69 cases were MVI positive in the training group, including 45 cases in the low-risk group (M1) and 24 cases in the high-risk group (M2), and 47 cases were MVI negative (M0). In the training group, the optimal subset features were obtained through feature selection, and the arterial phase radiomics model, portal venous phase radiomics model, delayed phase radiomics model, three-phase radiomics model, clinical imaging model, and combined model were developed using Linear Support Vector Classification. The test group was used for validation, and the efficacy of each model was evaluated through the receiver operating characteristic curve (ROC).

RESULTS

The clinical imaging features of MVI included alpha-fetoprotein, tumor size, tumor margin, peritumoral enhancement, intratumoral artery, and low-density halo. The area under the curve (AUC) of the ROC values of the clinical imaging model for M0, M1, and M2 were 0.831, 0.701, and 0.847, respectively, in the training group and 0.782, 0.534, and 0.785, respectively, in the test group. After combined radiomics analyis, the AUC values for M0, M1, and M2 in the test group were 0.818, 0.688, and 0.867, respectively. The difference between the clinical imaging model and the combined model was statistically significant (p = 0.029).

CONCLUSION

The clinical imaging model and radiomics model developed in this study had a specific predictive value for HCC MVI grading, which can provide precise reference value for preoperative clinical diagnosis and treatment. The combined application of the two models had a high predictive efficacy.

Radiobioinformatics: A novel bridge between basic research and clinical practice for clinical decision support in diffuse liver diseases

The liver is a multifaceted organ that is responsible for many critical functions encompassing amino acid, carbohydrate, and lipid metabolism, all of which make a healthy liver essential for the human body. Contemporary imaging methodologies have remarkable diagnostic accuracy in discerning focal liver lesions; however, a comprehensive understanding of diffuse liver diseases is a requisite for radiologists to accurately diagnose or predict the progression of such lesions within clinical contexts. Nonetheless, the conventional attributes of radiological features, including morphology, size, margin, density, signal intensity, and echoes, limit their clinical utility. Radiomics is a widely used approach that is characterized by the extraction of copious image features from radiographic depictions, which gives it considerable potential in addressing this limitation. It is worth noting that functional or molecular alterations occur significantly prior to the morphological shifts discernible by imaging modalities. Consequently, the explication of potential mechanisms by multiomics analyses (encompassing genomics, epigenomics, transcriptomics, proteomics, and metabolomics) is essential for investigating putative signal pathway regulations from a radiological viewpoint. In this review, we elaborate on the principal pathological categorizations of diffuse liver diseases, the evaluation of multiomics approaches pertaining to diffuse liver diseases, and the prospective value of predictive models. Accordingly, the overarching objective of this review is to scrutinize the interrelations between radiological features and bioinformatics as well as to consider the development of prediction models predicated on radiobioinformatics as integral components of clinical decision support systems for diffuse liver diseases.

Applying Lipidomics to Non-Alcoholic Fatty Liver Disease: A Clinical Perspective

The prevalence of Non-alcoholic fatty liver disease (NAFLD) and associated complications, such as hepatocellular carcinoma (HCC), is growing worldwide, due to the epidemics of metabolic risk factors, such as obesity and type II diabetes. Among other factors, an aberrant lipid metabolism represents a crucial step in the pathogenesis of NAFLD and the development of HCC in this population. In this review, we summarize the evidence supporting the application of translational lipidomics in NAFLD patients and NAFLD associated HCC in clinical practice.

Identification of Gut Microbial Lysine and Histidine Degradation and CYP-Dependent Metabolites as Biomarkers of Fatty Liver Disease

Numerous studies have described specific metabolites as biomarkers of severe liver diseases, but very few have measured gut microbiota (GM)-produced metabolites in fatty liver disease. We aimed at finding GM signatures and metabolite markers in plasma and feces related to high liver fat content. Based on imaging, we divided study participants into low (<5%, LF, n = 25) and high (>5%, HF, n = 39) liver fat groups. Fecal (LF n = 14, HF n = 25) and plasma (LF n = 11, HF n = 7) metabolomes of subsets of participants were studied using liquid chromatography/high resolution mass spectrometry. The GM were analyzed using 16S rRNA gene sequencing. Additionally, blood clinical variables and diet were studied. Dyslipidemia, higher liver enzymes and insulin resistance characterized the HF group. No major differences in diet were found between the groups. In the GM, the HF group had lower abundance of Bacteroides and Prevotellaceae NK3B31 group than the LF group after adjusting for metformin use or obesity. In feces, the HF group had higher levels of lysine and histidine degradation products, while 6-hydroxybetatestosterone (metabolized by CYP3A4) was low. Higher plasma levels of caffeine and its metabolites in the HF group indicate that the activity of hepatic CYP1A2 was lower than in the LF group. Our results suggest, that low fecal Prevotellaceae NK3B31 and Bacteroides abundance, and increased lysine and histidine degradation may serve as GM biomarkers of high liver fat. Altered plasma caffeine metabolites and lowered testosterone metabolism may specify decreased CYP activities, and their potential utility, as biomarkers of fatty liver disease. IMPORTANCE Because the high prevalence of nonalcoholic fatty liver disease sets diagnostic challenges to health care, identification of new biomarkers of the disease that in the future could have potential utility as diagnostic biomarkers of high liver fat content is important. Our results show that increased amino acid degradation products in the feces may be such biomarkers. In the blood, molecules that indicate defective hepatic metabolic enzyme activities were identified in individuals with high liver fat content.

Non-invasive biomarkers for liver inflammation in non-alcoholic fatty liver disease: present and future

Inflammation is the key driver of liver fibrosis progression in non-alcoholic fatty liver disease (NAFLD). Unfortunately, it is often challenging to assess inflammation in NAFLD due to its dynamic nature and poor correlation with liver biochemical markers. Liver histology keeps its role as the standard tool, yet it is well-known for substantial sampling, intraobserver, and interobserver variability. Serum proinflammatory cytokines and apoptotic markers, namely cytokeratin-18, are well-studied with reasonable accuracy, whereas serum metabolomics and lipidomics have been adopted in some commercially available diagnostic models. Ultrasound and computed tomography imaging techniques are attractive due to their wide availability; yet their accuracies may not be comparable with magnetic resonance imaging-based tools. Machine learning and deep learning models, be they supervised or unsupervised learning, are promising tools to identify various subtypes of NAFLD, including those with dominating liver inflammation, contributing to sustainable care pathways for NAFLD.

Impairment of patient-reported outcomes among patients with non-alcoholic fatty liver disease: a registry-based study

BACKGROUND

Patients with non-alcoholic fatty liver disease (NAFLD) and more advanced fibrosis tend to have more impairment in their health-related quality of life and other patient-reported outcomes (PROs).

AIM

To assess the association of PROs with select non-invasive tests (NITs) for fibrosis including FAST, Agile 3+ and Agile 4 scores METHODS: We enrolled patients with an established diagnosis of NAFLD who were seen in a tertiary care clinic into the NAFLD/NASH Registry. The FAST, Agile 3+ and Agile 4 scores were calculated using liver stiffness measurements by transient elastography and laboratory parameters. PROs were assessed using FACIT-F, CLDQ-NASH and WPAI instruments (total of 17 domain and summary scores).

RESULTS

There were 1509 patients with NAFLD (mean age: 49 ± 11 years, 50% men, 41% employed, 30% advanced fibrosis and 20% cirrhosis). The mean FAST, Agile 3+ and Agile 4 scores were 0.39 ± 0.26, 0.35 ± 0.31 and 0.12 ± 0.23, respectively. Subjects with lower FAST, Agile 3+ and Agile 4 scores had the highest scores in select domains of FACIT-F, CLDQ-NASH and WPAI (p < 0.05 in comparison to subjects with elevated or high-risk NIT scores). Correlations with continuous NITs were significantly negative for Emotional and Functional well-being (FACIT-F), Activity/energy, Systemic symptoms, Worry and total scores (CLDQ-NASH), and Activity of WPAI (p < 0.05); the strongest was for Worry (CLDQ-NASH) with FAST (R = -0.17, p < 0.0001). The PRO scores of patients with NAFLD were lower than those of matched patients with chronic hepatitis B (p < 0.05 for 9/17 domain and summary scores).

CONCLUSION

Patients with NAFLD and high FAST, Agile 3+ or Agile 4 scores experience impairment of health-related quality of life.

Artificial intelligence models for the diagnosis and management of liver diseases

With the development of more advanced methods for the diagnosis and treatment of diseases, the data required for medical care are becoming complex, and misinterpretation of information due to human error may result in serious consequences. Human error can be avoided with the support of artificial intelligence (AI). AI models trained with various medical data for diagnosis and management of liver diseases have been applied to hepatitis, fatty liver disease, liver cirrhosis, and liver cancer. Some of these models have been reported to outperform human experts in terms of performance, indicating their potential for supporting clinical practice given their high-speed output. This paper summarizes the recent advances in AI for liver disease and introduces the AI-aided diagnosis of liver tumors using B-mode ultrasonography.

Detangling the interrelations between MAFLD, insulin resistance, and key hormones

Metabolic dysfunction-associated fatty liver disease (MAFLD) has increasingly become a significant and highly prevalent cause of chronic liver disease, displaying a wide array of risk factors and pathophysiologic mechanisms of which only a few have so far been clearly elucidated. A bidirectional interaction between hormonal discrepancies and metabolic-related disorders, including obesity, type 2 diabetes mellitus (T2DM), and polycystic ovarian syndrome (PCOS) has been described. Since the change in nomenclature from non-alcoholic fatty liver disease (NAFLD) to MAFLD is based on the clear impact of metabolic elements on the disease, the reciprocal interactions of hormones such as insulin, adipokines (leptin and adiponectin), and estrogens have strongly pointed to the intrinsic links that lead to the heterogeneous epidemiology, clinical presentations, and risk factors involved in MAFLD in different populations. The objective of this work is twofold. Firstly, there is a brief discussion regarding the change in nomenclature as well as epidemiology, risk factors, and pathophysiologic mechanisms other than hormonal effects, which include nutrition and the gut microbiome, as well as genetic and epigenetic influences. Secondly, we review the basis of the most important hormonal factors involved in the development and progression of MAFLD that act both independently and in an interrelated manner.

Applications of machine learning in metabolomics: Disease modeling and classification

Metabolomics research has recently gained popularity because it enables the study of biological traits at the biochemical level and, as a result, can directly reveal what occurs in a cell or a tissue based on health or disease status, complementing other omics such as genomics and transcriptomics. Like other high-throughput biological experiments, metabolomics produces vast volumes of complex data. The application of machine learning (ML) to analyze data, recognize patterns, and build models is expanding across multiple fields. In the same way, ML methods are utilized for the classification, regression, or clustering of highly complex metabolomic data. This review discusses how disease modeling and diagnosis can be enhanced via deep and comprehensive metabolomic profiling using ML. We discuss the general layout of a metabolic workflow and the fundamental ML techniques used to analyze metabolomic data, including support vector machines (SVM), decision trees, random forests (RF), neural networks (NN), and deep learning (DL). Finally, we present the advantages and disadvantages of various ML methods and provide suggestions for different metabolic data analysis scenarios.

Artificial intelligence applied to omics data in liver diseases: Enhancing clinical predictions

Rapid development of biotechnology has led to the generation of vast amounts of multi-omics data, necessitating the advancement of bioinformatics and artificial intelligence to enable computational modeling to diagnose and predict clinical outcome. Both conventional machine learning and new deep learning algorithms screen existing data unbiasedly to uncover patterns and create models that can be valuable in informing clinical decisions. We summarized published literature on the use of AI models trained on omics datasets, with and without clinical data, to diagnose, risk-stratify, and predict survivability of patients with non-malignant liver diseases. A total of 20 different models were tested in selected studies. Generally, the addition of omics data to regular clinical parameters or individual biomarkers improved the AI model performance. For instance, using NAFLD fibrosis score to distinguish F0-F2 from F3-F4 fibrotic stages, the area under the curve (AUC) was 0.87. When integrating metabolomic data by a GMLVQ model, the AUC drastically improved to 0.99. The use of RF on multi-omics and clinical data in another study to predict progression of NAFLD to NASH resulted in an AUC of 0.84, compared to 0.82 when using clinical data only. A comparison of RF, SVM and kNN models on genomics data to classify immune tolerant phase in chronic hepatitis B resulted in AUC of 0.8793-0.8838 compared to 0.6759-0.7276 when using various serum biomarkers. Overall, the integration of omics was shown to improve prediction performance compared to models built only on clinical parameters, indicating a potential use for personalized medicine in clinical setting.

A machine-learning approach for nonalcoholic steatohepatitis susceptibility estimation

BACKGROUND

Nonalcoholic steatohepatitis (NASH), a severe form of nonalcoholic fatty liver disease, can lead to advanced liver damage and has become an increasingly prominent health problem worldwide. Predictive models for early identification of high-risk individuals could help identify preventive and interventional measures. Traditional epidemiological models with limited predictive power are based on statistical analysis. In the current study, a novel machine-learning approach was developed for individual NASH susceptibility prediction using candidate single nucleotide polymorphisms (SNPs).

METHODS

A total of 245 NASH patients and 120 healthy individuals were included in the study. Single nucleotide polymorphism genotypes of candidate genes including two SNPs in the cytochrome P450 family 2 subfamily E member 1 (CYP2E1) gene (rs6413432, rs3813867), two SNPs in the glucokinase regulator (GCKR) gene (rs780094, rs1260326), rs738409 SNP in patatin-like phospholipase domain-containing 3 (PNPLA3), and gender parameters were used to develop models for identifying at-risk individuals. To predict the individual's susceptibility to NASH, nine different machine-learning models were constructed. These models involved two different feature selections including Chi-square, and support vector machine recursive feature elimination (SVM-RFE) and three classification algorithms including k-nearest neighbor (KNN), multi-layer perceptron (MLP), and random forest (RF). All nine machine-learning models were trained using 80% of both the NASH patients and the healthy controls data. The nine machine-learning models were then tested on 20% of both groups. The model's performance was compared for model accuracy, precision, sensitivity, and F measure.

RESULTS

Among all nine machine-learning models, the KNN classifier with all features as input showed the highest performance with 86% F measure and 79% accuracy.

CONCLUSIONS

Machine learning based on genomic variety may be applicable for estimating an individual's susceptibility for developing NASH among high-risk groups with a high degree of accuracy, precision, and sensitivity.

Menopausal hormone therapy in women with dyslipidemia and nonalcoholic fatty liver disease

The cessation of ovarian function is associated with an increase in abdominal adipose tissue, dyslipidemia, and nonalcoholic fatty liver disease (NAFLD), which may contribute to the augmented cardiovascular risk observed in postmenopausal women. After ovarian function stops, circulating triglyceride, total cholesterol, and low-density lipoprotein-cholesterol (LDL-C) concentrations increase, whereas high-density lipoprotein-cholesterol (HDL-C) and lipoprotein (Lp(a)) remain essentially unchanged. Similarly, the rates of NAFLD, possibly including the advanced forms of the disease (e.g., hepatic fibrosis), increase in postmenopausal compared with premenopausal women. These effects make menopausal hormone therapy (MHT) an attractive way to restore them. Estrogen per os decreases LDL-C and Lp(a) and increases HDL-C and triglyceride concentrations. The transdermal administration of estrogen has a more neutral effect on triglycerides, albeit a less beneficial effect on LDL-C, HDL-C, and Lp(a). Co-administration of a progestagen diminishes the effect of estrogen on LDL-C, HDL-C, and Lp(a), which, however, remains beneficial. Importantly, the effect may vary with different progestagens, being lesser with natural progesterone and dydrogesterone. Regarding the effect of MHT on NAFLD, though experimental data are currently favorable, clinical evidence is to date limited and controversial. Therefore, there is a need for specifically designed clinical trials, ideally with paired liver biopsies, to demonstrate the effect of different MHT schemes on NAFLD, which is of considerable importance, given that NAFLD is more prevalent after the cessation of ovarian function.

Advance of Serum Biomarkers and Combined Diagnostic Panels in Nonalcoholic Fatty Liver Disease

Nonalcoholic fatty liver disease (NAFLD) affects approximately 25-30% population worldwide, which progresses from simple steatosis to nonalcoholic steatohepatitis (NASH), fibrosis, cirrhosis, and hepatocellular carcinoma, and has complications such as cardiovascular events. Liver biopsy is still the gold standard for the diagnosis of NAFLD, with some limitations, such as invasive, sampling deviation, and empirical judgment. Therefore, it is urgent to develop noninvasive diagnostic biomarkers. Currently, a large number of NAFLD-related serum biomarkers have been identified, including apoptosis, inflammation, fibrosis, adipokines, hepatokines, and omics biomarkers, which could effectively diagnose NASH and exclude patients with progressive fibrosis. We summarized serum biomarkers and combined diagnostic panels of NAFLD, to provide some guidance for the noninvasive diagnosis and further clinical studies.

Applications of artificial intelligence (AI) in researches on non-alcoholic fatty liver disease(NAFLD) : A systematic review

Non-alcoholic fatty liver disease (NAFLD) is one of the most important causes of chronic liver disease in the world, it has been found that cardiovascular and renal risks and diseases are also highly prevalent in adults with NAFLD. Diagnosis and treatment of NAFLD face many challenges, although the medical science has been very developed. Efficiency, accuracy and individualization are the main goals to be solved. Evaluation of the severity of NAFLD involves a variety of clinical parameters, how to optimize non-invasive evaluation methods is a necessary issue that needs to be discussed in this field. Artificial intelligence (AI) has become increasingly widespread in healthcare applications, and it has been also brought many new insights into better analyzing chronic liver disease, including NAFLD. This paper reviewed AI related researches in NAFLD field published recently, summarized diagnostic models based on electronic health record and lab test, ultrasound and radio imaging, and liver histopathological data, described the application of therapeutic models in personalized lifestyle guidance and the development of drugs for NAFLD. In addition, we also analyzed present AI models in distinguishing healthy VS NAFLD/NASH, and fibrosis VS non-fibrosis in the evaluation of NAFLD progression. We hope to provide alternative directions for the future research.

Metabolomics in Bariatric and Metabolic Surgery Research and the Potential of Deep Learning in Bridging the Gap

During the past several years, there has been a shift in terminology from bariatric surgery alone to bariatric and metabolic surgery (BMS). More than a change in name, this signifies a paradigm shift that incorporates the metabolic effects of operations performed for weight loss and the amelioration of related medical problems. Metabolomics is a relatively novel concept in the field of bariatrics, with some consistent changes in metabolite concentrations before and after weight loss. However, the abundance of metabolites is not easy to handle. This is where artificial intelligence, and more specifically deep learning, would aid in revealing hidden relationships and would help the clinician in the decision-making process of patient selection in an individualized way.

Application of artificial intelligence in non-alcoholic fatty liver disease and viral hepatitis

Non-alcoholic fatty liver disease (NAFLD) and chronic viral hepatitis are among the most significant causes of liver-related mortality worldwide. It is critical to develop reliable methods of predicting progression to fibrosis, cirrhosis, and decompensated liver disease. Current screening methods such as biopsy and transient elastography are limited by invasiveness and observer variation in analysis of data. Artificial intelligence (AI) provides a unique opportunity to more accurately diagnose NAFLD and viral hepatitis, and to identify patients at high risk for disease progression. We conducted a literature review of existing evidence for AI in NAFLD and viral hepatitis. Thirteen articles on AI in NAFLD and 14 on viral hepatitis were included in our analysis. We found that machine learning algorithms were comparable in accuracy to current methods for diagnosis and fibrosis prediction (MELD-Na score, liver biopsy, FIB-4 score, and biomarkers). They also reliably predicted hepatitis C treatment failure and hepatic encephalopathy, for which there are currently no established prediction tools. These studies show that AI could be a helpful adjunct to existing techniques for diagnosing, monitoring, and treating both NAFLD and viral hepatitis.

Nutrients handling after bariatric surgery, the role of gastrointestinal adaptation

Bariatric surgery determines a rearrangement of the gastrointestinal tract that influences nutrient handling and plays a role in the metabolic changes observed after surgery. Most of the changes depend on the accelerated gastric emptying observed in Roux-en-Y gastric bypass (RYGB) and, to a lesser extent, in sleeve gastrectomy (SG). The rapid delivery of meal into the jejunum, particularly after RYGB, contributes to the prompt appearance of glucose in peripheral circulation. Glucose increase is the principal determinant of GLP-1 increase with the consequent stimulation of insulin secretion, the latter balanced by a paradoxical glucagon increase that stimulates EGP to prevent hypoglycaemia. Protein digestion and amino acid absorption appear accelerated after RYGB but not after SG. After RYGB, the adaptation of the gut to the new condition participates to the metabolic change. The intestinal transit is delayed, the gut microbioma is changed, the epithelium becomes hypertrophic and increases the expression of glucose transporter and of the number of cell secreting hormones. These changes are not observed after SG. After RYGB-less after SG-bile acids (BA) increase, influencing glucose metabolism probably modulating FXR and TGR5 with an effect on insulin sensitivity. Muscle, hepatic and adipose tissue insulin sensitivity improve, and the gut reinforces the recovery of IS by enhancing glucose uptake and through the effect of the BA. The intestinal changes observed after RYGB result in a light malabsorption of lipid but not of carbohydrate and protein. In conclusion, functional and morphological adaptations of the gut after RYGB and SG activate inter-organs cross-talk that modulates the metabolic changes observed after surgery.Level of evidence Level V, narrative literature review.

Machine learning to predict progression of non-alcoholic fatty liver to non-alcoholic steatohepatitis or fibrosis

Background

Non-alcoholic fatty liver (NAFL) can progress to the severe subtype non-alcoholic steatohepatitis (NASH) and/or fibrosis, which are associated with increased morbidity, mortality, and healthcare costs. Current machine learning studies detect NASH; however, this study is unique in predicting the progression of NAFL patients to NASH or fibrosis.

Aim

To utilize clinical information from NAFL-diagnosed patients to predict the likelihood of progression to NASH or fibrosis.

Methods

Data were collected from electronic health records of patients receiving a first-time NAFL diagnosis. A gradient boosted machine learning algorithm (XGBoost) as well as logistic regression (LR) and multi-layer perceptron (MLP) models were developed. A five-fold cross-validation grid search was utilized for hyperparameter optimization of variables, including maximum tree depth, learning rate, and number of estimators. Predictions of patients likely to progress to NASH or fibrosis within 4 years of initial NAFL diagnosis were made using demographic features, vital signs, and laboratory measurements.

Results

The XGBoost algorithm achieved area under the receiver operating characteristic (AUROC) values of 0.79 for prediction of progression to NASH and 0.87 for fibrosis on both hold-out and external validation test sets. The XGBoost algorithm outperformed the LR and MLP models for both NASH and fibrosis prediction on all metrics.

Conclusion

It is possible to accurately identify newly diagnosed NAFL patients at high risk of progression to NASH or fibrosis. Early identification of these patients may allow for increased clinical monitoring, more aggressive preventative measures to slow the progression of NAFL and fibrosis, and efficient clinical trial enrollment.

Machine learning using longitudinal prescription and medical claims for the detection of non-alcoholic steatohepatitis (NASH)

Objectives

To develop and evaluate machine learning models to detect patients with suspected undiagnosed non-alcoholic steatohepatitis (NASH) for diagnostic screening and clinical management.

Methods

In this retrospective observational non-interventional study using administrative medical claims data from 1 463 089 patients, gradient-boosted decision trees were trained to detect patients with likely NASH from an at-risk patient population with a history of obesity, type 2 diabetes mellitus, metabolic disorder or non-alcoholic fatty liver (NAFL). Models were trained to detect likely NASH in all at-risk patients or in the subset without a prior NAFL diagnosis (at-risk non-NAFL patients). Models were trained and validated using retrospective medical claims data and assessed using area under precision recall curves and receiver operating characteristic curves (AUPRCs and AUROCs).

Results

The 6-month incidences of NASH in claims data were 1 per 1437 at-risk patients and 1 per 2127 at-risk non-NAFL patients. The model trained to detect NASH in all at-risk patients had an AUPRC of 0.0107 (95% CI 0.0104 to 0.0110) and an AUROC of 0.84. At 10% recall, model precision was 4.3%, which is 60× above NASH incidence. The model trained to detect NASH in the non-NAFL cohort had an AUPRC of 0.0030 (95% CI 0.0029 to 0.0031) and an AUROC of 0.78. At 10% recall, model precision was 1%, which is 20× above NASH incidence.

Conclusion

The low incidence of NASH in medical claims data corroborates the pattern of NASH underdiagnosis in clinical practice. Claims-based machine learning could facilitate the detection of patients with probable NASH for diagnostic testing and disease management.

The application of artificial intelligence in hepatology: A systematic review

The integration of human and artificial intelligence (AI) in medicine has only recently begun but it has already become obvious that intelligent systems can dramatically improve the management of liver diseases. Big data made it possible to envisage transformative developments of the use of AI for diagnosing, predicting prognosis and treating liver diseases, but there is still a lot of work to do. If we want to achieve the 21^st century digital revolution, there is an urgent need for specific national and international rules, and to adhere to bioethical parameters when collecting data. Avoiding misleading results is essential for the effective use of AI. A crucial question is whether it is possible to sustain, technically and morally, the process of integration between man and machine. We present a systematic review on the applications of AI to hepatology, highlighting the current challenges and crucial issues related to the use of such technologies.

Diagnostic Modalities of Non-Alcoholic Fatty Liver Disease: From Biochemical Biomarkers to Multi-Omics Non-Invasive Approaches

Non-Alcoholic Fatty Liver Disease (NAFLD) is currently the most common cause of chronic liver disease worldwide, and its prevalence is increasing globally. NAFLD is a multifaceted disorder, and its spectrum includes steatosis to steatohepatitis, which may evolve to advanced fibrosis and cirrhosis. In addition, the presence of NAFLD is independently associated with a higher cardiometabolic risk and increased mortality rates. Considering that the vast majority of individuals with NAFLD are mainly asymptomatic, early diagnosis of non-alcoholic steatohepatitis (NASH) and accurate staging of fibrosis risk is crucial for better stratification, monitoring and targeted management of patients at risk. To date, liver biopsy remains the gold standard procedure for the diagnosis of NASH and staging of NAFLD. However, due to its invasive nature, research on non-invasive tests is rapidly increasing with significant advances having been achieved during the last decades in the diagnostic field. New promising non-invasive biomarkers and techniques have been developed, evaluated and assessed, including biochemical markers, imaging modalities and the most recent multi-omics approaches. Our article provides a comprehensive review of the currently available and emerging non-invasive diagnostic tools used in assessing NAFLD, also highlighting the importance of accurate and validated diagnostic tools.

Artificial Intelligence in Hepatology: A Narrative Review

There has been a tremendous growth in data collection in hepatology over the last decade. This wealth of "big data" lends itself to the application of artificial intelligence in the development of predictive and diagnostic models with potentially greater accuracy than standard biostatistics. As processing power of computing systems has improved and data are made more accessible through the large databases and electronic health record, these more contemporary techniques for analyzing and interpreting data have garnered much interest in the field of medicine. This review highlights the current evidence base for the use of artificial intelligence in hepatology, focusing particularly on the areas of diagnosis and prognosis of advanced chronic liver disease and hepatic neoplasia.

Artificial intelligence in gastroenterology: A state-of-the-art review

The development of artificial intelligence (AI) has increased dramatically in the last 20 years, with clinical applications progressively being explored for most of the medical specialties. The field of gastroenterology and hepatology, substantially reliant on vast amounts of imaging studies, is not an exception. The clinical applications of AI systems in this field include the identification of premalignant or malignant lesions (e.g., identification of dysplasia or esophageal adenocarcinoma in Barrett’s esophagus, pancreatic malignancies), detection of lesions (e.g., polyp identification and classification, small-bowel bleeding lesion on capsule endoscopy, pancreatic cystic lesions), development of objective scoring systems for risk stratification, predicting disease prognosis or treatment response [e.g., determining survival in patients post-resection of hepatocellular carcinoma), determining which patients with inflammatory bowel disease (IBD) will benefit from biologic therapy], or evaluation of metrics such as bowel preparation score or quality of endoscopic examination. The objective of this comprehensive review is to analyze the available AI-related studies pertaining to the entirety of the gastrointestinal tract, including the upper, middle and lower tracts; IBD; the hepatobiliary system; and the pancreas, discussing the findings and clinical applications, as well as outlining the current limitations and future directions in this field.

Defining NASH from a Multi-Omics Systems Biology Perspective

Non-alcoholic steatohepatitis (NASH) is a chronic liver disease affecting up to 6.5% of the general population. There is no simple definition of NASH, and the molecular mechanism underlying disease pathogenesis remains elusive. Studies applying single omics technologies have enabled a better understanding of the molecular profiles associated with steatosis and hepatic inflammation—the commonly accepted histologic features for diagnosing NASH, as well as the discovery of novel candidate biomarkers. Multi-omics analysis holds great potential to uncover new insights into disease mechanism through integrating multiple layers of molecular information. Despite the technical and computational challenges associated with such efforts, a few pioneering studies have successfully applied multi-omics technologies to investigate NASH. Here, we review the most recent technological developments in mass spectrometry (MS)-based proteomics, metabolomics, and lipidomics. We summarize multi-omics studies and emerging omics biomarkers in NASH and highlight the biological insights gained through these integrated analyses.

Serum lipids are associated with nonalcoholic fatty liver disease: a pilot case-control study in Mexico

Background

Nonalcoholic fatty liver disease (NAFLD) is a leading cause of chronic liver disease and cirrhosis. NAFLD is mediated by changes in lipid metabolism and known risk factors include obesity, metabolic syndrome, and diabetes. The aim of this study was to better understand differences in the lipid composition of individuals with NAFLD compared to controls, by performing direct infusion lipidomics on serum biospecimens from a cohort study of adults in Mexico.

Methods

A nested case-control study was conducted with a sample of 98 NAFLD cases and 100 healthy controls who are participating in an on-going, longitudinal study in Mexico. NAFLD cases were clinically confirmed using elevated liver enzyme tests and liver ultrasound or liver ultrasound elastography, after excluding alcohol abuse, and 100 controls were identified as having at least two consecutive normal alanine aminotransferase (ALT) and aspartate aminotransferase (AST) (< 40 U/L) results in a 6-month period, and a normal liver ultrasound elastography result in January 2018. Samples were analyzed on the Sciex Lipidyzer Platform and quantified with normalization to serum volume. As many as 1100 lipid species can be identified using the Lipidyzer targeted multiple-reaction monitoring list. The association between serum lipids and NAFLD was investigated using analysis of covariance, random forest analysis, and by generating receiver operator characteristic (ROC) curves.

Results

NAFLD cases had differences in total amounts of serum cholesterol esters, lysophosphatidylcholines, sphingomyelins, and triacylglycerols (TAGs), however, other lipid subclasses were similar to controls. Analysis of individual TAG species revealed increased incorporation of saturated fatty acyl tails in serum of NAFLD cases. After adjusting for age, sex, body mass index, and PNPLA3 genotype, a combined panel of ten lipids predicted case or control status better than an area under the ROC curve of 0.83.

Conclusions

These preliminary results indicate that the serum lipidome differs in patients with NAFLD, compared to healthy controls, and suggest that assessing the desaturation state of TAGs or a specific lipid panel may be useful clinical tools for the diagnosis of NAFLD.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12944-021-01526-5.

LC-MS-based lipidomic analysis in distinguishing patients with nonalcoholic steatohepatitis from nonalcoholic fatty liver

BACKGROUND

Nonalcoholic fatty liver disease (NAFLD) is one of the main liver diseases, and its pathologic profile includes nonalcoholic fatty liver (NAFL) and nonalcoholic steatohepatitis (NASH). However, there is no reliable non-invasive parameter in distinguishing NASH from NAFL in clinical practice. The present study was to find a non-invasive way to differentiate these two categories of NAFLD via lipidomic analysis.

METHODS

Lipidomic analysis was used to determine the changes of lipid moieties in blood from 20 NAFL and 10 NASH patients with liver biopsy. Liver histology was evaluated after hematoxylin and eosin staining and Masson's trichrome staining. The profile of lipid metabolites in correlation with steatosis, inflammation, hepatocellular necroptosis, fibrosis, and NAFLD activity score (NAS) was analyzed.

RESULTS

Compared with NAFL patients, NASH patients had higher degree of steatosis, ballooning degeneration, lobular inflammation. A total of 434 different lipid molecules were identified, which were mainly composed of various phospholipids and triacylglycerols. Many lipids, such as phosphatidylcholine (PC) (P-22:0/18:1), sphingomyelin (SM) (d14:0/18:0), SM (d14:0/24:0), SM (d14:0/22:0), phosphatidylethanolamine (PE) (18:0/22:5), PC (O-22:2/12:0), and PC (26:1/11:0) were elevated in the NASH group compared to those in the NAFL group. Specific analysis revealed an overall lipidomic profile shift from NAFL to NASH, and identified valuable lipid moieties, such as PCs [PC (14:0/18:2), PE (18:0/22:5) and PC (26:1/11:0)] or plasmalogens [PC (O-22:0/0:0), PC (O-18:0/0:0), PC (O-16:0/0:0)], which were significantly altered in NASH patients. In addition, PC (14:0/18:2), phosphatidic acid (18:2/24:4) were positively correlated with NAS; whereas PC (18:0/0:0) was correlated positively with fibrosis score.

CONCLUSIONS

The present study revealed overall lipidomic profile shift from NAFL to NASH, identified valuable lipid moieties which may be non-invasive biomarkers in the categorization of NAFLD. The correlations between lipid moieties and NAS and fibrosis scores indicate that these lipid biomarkers may be used to predict the severity of the disease.

Genetic risk scores and personalization of care in fatty liver disease

Nonalcoholic fatty liver disease is the leading cause of chronic liver disease. Genetic predisposition, lifestyle, and metabolic comorbidities concur to nonalcoholic fatty liver disease development and progression to nonalcoholic steatohepatitis, cirrhosis, and hepatocellular carcinoma. Improvement in risk stratification and development of effective therapies for nonalcoholic steatohepatitis are key unmet clinical needs. Knowledge emerging from genomics could meet this need. A polygenic risk score (PRS) is calculated by summing the number of trait-associated alleles carried by an individual, which can be weighted by their effect size on the trait and captures the individual's genetic risk to develop a disease. In this review, we focalize on the potential use of PRSs for disease detection at an early stage and stratification of the risk of progression to severe forms. PRSs may represent robust instruments to implement targeted prevention programs, hepatocellular carcinoma surveillance in at-risk individuals, and to develop precision medicine therapeutic approaches.