Purpose: Oxidative stress has been implicated in several disorders, including acute pancreatitis (AP). Ischemia-modified albumin (IMA), which reflects the ability to bind cobalt, has been found to be elevated in conditions of oxidative stress and tissue hypoxia. This study examined IMA and adjusted IMA levels in patients with AP, and examined the associations of IMA and adjusted IMA levels to the severity of AP. Patients and methods: A total of 42 consecutive patients with AP and 43 age- and sex-matched control subjects were enrolled. Serum samples were obtained from patients with AP on admission as well as 48–72 hours after hospitalization, and from the controls, at the time of enrollment. Adjusted IMA was calculated by multiplying the IMA value of each patient with the ratio of the patient’s albumin value and the median albumin value of the study population. The severity of AP was assessed according to the modified Atlanta classification, and the patients were divided into 2 groups: mild AP and severe AP. Results: The serum IMA and adjusted IMA values of patients with AP on admission and those of the controls did not differ (p=0.86 and p=0.99, respectively). The second measurements of IMA and adjusted IMA in the AP group were higher than the first measurements of both the AP group and controls (for all, p<0.01). Among the IMA measurements, only adjusted IMA on admission had the ability to predict the severity of AP. Severe AP was correlated with albumin, and the area under the curve of adjusted IMA values on admission was 0.746 for differentiating patients with severe AP from mild AP with statistical significance (p=0.005). Conclusion: It was shown that IMA and adjusted IMA levels rise with the progression of AP. Lower levels of adjusted IMA predict the severity of AP. Further studies with serial measurements of IMA are warranted to explore the indicative role of IMA in the course of AP.

Chronic viral hepatitis affects nearly one half billion people; it can result in hepatic fibrosis and cirrhosis if untreated. We investigated whether the polypeptides, adropin, apelin, elabela, asprosin and betatrophin, could be useful biomarkers for diagnosing chronic hepatitis and for the staging fibrosis. Patients 18−60 years old who underwent a liver biopsy for chronic hepatitis B (CHB) or C (CHC) from January 2014 to January 2019 were included in our study. The patients were divided into three groups: control group, CHB group, and CHC group. The CHB group comprised four subgroups based on the severity of the hepatic fibrosis. Liver biopsy specimens of all groups were evaluated for adropin, apelin, elabela, asprosin and betatrophin immunoreactivity by light microscopy. Adropin, apelin, elabela and betatrophin were immunoreactive in the hepatocytes, while asprosin was not in any group. In the CHB group, adropin and elabela immunoreactivity was increased significantly in stages III and IV patients compared to the other subgroups, whereas stages I and II patients were comparable to the control group. The CHC group exhibited decreased betatrophin immunoreactivity and increased elabela immunoreactivity compared to the control group. We suggest that adropin and elabela can provide clues for staging and monitoring fibrosis in CHB, and may be potentially useful biomarkers.

Background: Intestinal metaplasia (IM) is a precancerous gastric condition that requires accurate histopathological diagnosis to enable early intervention and cancer prevention. Traditional evaluation of H&E-stained tissue slides can be labor-intensive and prone to interobserver variability. Recent advances in deep learning, particularly transformer-based models, offer promising tools for improving diagnostic accuracy. Methods: We propose ViSwNeXtNet, a novel patch-wise ensemble framework that integrates three transformer-based architectures—ConvNeXt-Tiny, Swin-Tiny, and ViT-Base—for deep feature extraction. Features from each model (12,288 per model) were concatenated into a 36,864-dimensional vector and refined using iterative neighborhood component analysis (INCA) to select the most discriminative 565 features. A quadratic SVM classifier was trained using these selected features. The model was evaluated on two datasets: (1) a custom-collected dataset consisting of 516 intestinal metaplasia cases and 521 control cases, and (2) the public GasHisSDB dataset, which includes 20,160 normal and 13,124 abnormal H&E-stained image patches of size 160 × 160 pixels. Results: On the collected dataset, the proposed method achieved 94.41% accuracy, 94.63% sensitivity, and 94.40% F1 score. On the GasHisSDB dataset, it reached 99.20% accuracy, 99.39% sensitivity, and 99.16% F1 score, outperforming individual backbone models and demonstrating strong generalizability across datasets. Conclusions: ViSwNeXtNet successfully combines local, regional, and global representations of tissue structure through an ensemble of transformer-based models. The addition of INCA-based feature selection significantly enhances classification performance while reducing dimensionality. These findings suggest the method’s potential for integration into clinical pathology workflows. Future work will focus on multiclass classification, multicenter validation, and integration of explainable AI techniques.