Development and validation of a radiomics signature as a non-invasive complementary predictor of gastroesophageal varices and high-risk varices in compensated advanced chronic liver disease: A multicenter study.

BACKGROUND AND AIM: Gastroesophageal varices (GEV) present in compensated advanced chronic liver disease (cACLD) and can develop into high-risk varices (HRV). The gold standard for diagnosing GEV is esophagogastroduodenoscopy (EGD). However, EGD is invasive and less tolerant. This study aimed to develop and validate radiomics signatures based on noncontrast-enhanced computed tomography (CT) images for non-invasive diagnosis of GEV and HRV in patients with cACLD. METHODS: The multicenter trial enrolled 161 patients with cACLD from six university hospitals in China between January 2015 and September 2019, who underwent both EGD and noncontrast-enhanced CT examination within 14 days prior to the endoscopy. Two radiomics signatures, termed rGEV and rHRV, respectively, were built based on CT images in a training cohort of 129 patients and validated in a prospective validation cohort of 32 patients (ClinicalTrials. gov identifier: NCT03749954). RESULTS: In the training cohort, both rGEV and rHRV exhibited high discriminative abilities on determining the existence of GEV and HRV with the area under receiver operating characteristic curve (AUC) of 0.941 (95% confidence interval [CI] 0.904-0.978) and 0.836 (95% CI 0.766-0.905), respectively. In validation cohort, rGEV and rHRV showed high discriminative abilities with AUCs of 0.871 (95% CI 0.739-1.000) and 0.831 (95% CI 0.685-0.978), respectively. CONCLUSIONS: This study demonstrated that rGEV and rHRV could serve as the satisfying auxiliary parameters for detection of GEV and HRV with good diagnostic performance.

Different ratios of DHA/EPA reverses insulin resistance by improving adipocyte dysfunction and lipid disorders in HFD-induced IR mice.

Objective: Insulin resistance (IR) is linked to the development of diabetes, non-alcoholic fatty liver disease (NAFLD), and cardiovascular disease (CVDs). Docosahexaenoic acid (DHA) and eicosapentaenoic acid (EPA) from fish oils (FOs) were used to investigate their potential in high-fat diet (HFD)-induced IR mice under different ratios. Methods: A total of 84 male C57BL/6J (6 weeks old) mice were fed with HFD containing 45% kcal from fat for 16 weeks to establish the IR model. The IR mice were then fed with HFD or HFD + 4% DHA/EPA with different ratios (3 : 1, 1.5 : 1, 1 : 1, 1 : 1.5, 1 : 3, respectively) for another 12 weeks. During the experiment, the CON group (n = 12) was set to feed with a basic diet containing 10% kcal from fat. Results: HFD feeding for 16 weeks reduced insulin sensitivity and accelerated hypertrophy of white adipose tissue (WAT). Different ratios of DHA/EPA except for 1 : 1 decreased the HOMA-IR index, average area of adipocytes, and serum MDA, but increased the protein expression of PI3K. All ratios of DHA/EPA increased the protein expression of IRS-1, GLUT4, and adiponectin. Moreover, dietary DHA/EPA changed serum fatty acid (FA) composition by increasing the serum concentration of n-3 PUFAs. DHA/EPA supplements also improved serum lipid profiles (TG/TC/LDL-c/HDL-c, FFA) and reduced the hepatic steatosis area. Conclusions: The results indicate that an appropriate higher ratio of DHA (1.5 : 1) in DHA/EPA supplementation is recommended for IR prevention.

Machine learning-based CT radiomics method for predicting hospital stay in patients with pneumonia associated with SARS-CoV-2 infection: a multicenter study.

BACKGROUND: The coronavirus disease 2019 (COVID-19) has become a global challenge since the December 2019. The hospital stay is one of the prognostic indicators, and its predicting model based on CT radiomics features is important for assessing the patients' clinical outcome. The study aimed to develop and test machine learning-based CT radiomics models for predicting hospital stay in patients with COVID-19 pneumonia. METHODS: This retrospective, multicenter study enrolled patients with laboratory-confirmed SARS-CoV-2 infection and their initial CT images from 5 designated hospitals in Ankang, Lishui, Lanzhou, Linxia, and Zhenjiang between January 23, 2020 and February 8, 2020. Patients were classified into short-term (≤10 days) and long-term hospital stay (>10 days). CT radiomics models based on logistic regression (LR) and random forest (RF) were developed on features from pneumonia lesions in first four centers. The predictive performance was evaluated in fifth center (test dataset) on lung lobe- and patients-level. RESULTS: A total of 52 patients were enrolled from designated hospitals. As of February 20, 21 patients remained in hospital or with non-findings in CT were excluded. Therefore, 31 patients with 72 lesion segments were included in analysis. The CT radiomics models based on 6 second-order features were effective in discriminating short- and long-term hospital stay in patients with COVID-19 pneumonia, with areas under the curves of 0.97 (95% CI, 0.83-1.0) and 0.92 (95% CI, 0.67-1.0) by LR and RF, respectively, in test. The LR and RF model showed a sensitivity and specificity of 1.0 and 0.89, 0.75 and 1.0 in test respectively. As of February 28, a prospective cohort of six discharged patients were all correctly recognized as long-term stay using RF and LR models. CONCLUSIONS: The machine learning-based CT radiomics features and models showed feasibility and accuracy for predicting hospital stay in patients with COVID-19 pneumonia.