Artificial intelligence-assisted system for the assessment of Forrest classification of peptic ulcer bleeding: a multicenter diagnostic study.

BACKGROUND: Inaccurate Forrest classification may significantly affect clinical outcomes, especially in high risk patients. Therefore, this study aimed to develop a real-time deep convolutional neural network (DCNN) system to assess the Forrest classification of peptic ulcer bleeding (PUB). METHODS: A training dataset (3868 endoscopic images) and an internal validation dataset (834 images) were retrospectively collected from the 900th Hospital, Fuzhou, China. In addition, 521 images collected from four other hospitals were used for external validation. Finally, 46 endoscopic videos were prospectively collected to assess the real-time diagnostic performance of the DCNN system, whose diagnostic performance was also prospectively compared with that of three senior and three junior endoscopists. RESULTS: The DCNN system had a satisfactory diagnostic performance in the assessment of Forrest classification, with an accuracy of 91.2% (95%CI 89.5%-92.6%) and a macro-average area under the receiver operating characteristic curve of 0.80 in the validation dataset. Moreover, the DCNN system could judge suspicious regions automatically using Forrest classification in real-time videos, with an accuracy of 92.0% (95%CI 80.8%-97.8%). The DCNN system showed more accurate and stable diagnostic performance than endoscopists in the prospective clinical comparison test. This system helped to slightly improve the diagnostic performance of senior endoscopists and considerably enhance that of junior endoscopists. CONCLUSION: The DCNN system for the assessment of the Forrest classification of PUB showed satisfactory diagnostic performance, which was slightly superior to that of senior endoscopists. It could therefore effectively assist junior endoscopists in making such diagnoses during gastroscopy.

Phosphorylation of eIF2α mitigates endoplasmic reticulum stress and hepatocyte necroptosis in acute liver injury.

INTRODUCTION AND OBJECTIVES: Necroptosis and endoplasmic reticulum (ER) stress has been implicated in acute and chronic liver injury. Activated eukaryotic initiation factor 2 alpha (eIF2α) attenuates protein synthesis and relieves the load of protein folding in the ER. In this study, we aimed to analyze the impact of eIF2α phosphorylation on hepatocyte necroptosis in acute liver injury. MATERIALS AND METHODS: Male BALB/c mice were injected with tunicamycin or d-galactosamine, and LO2 cells were incubated with tunicamycin to induce acute liver injury. 4-Phenylbutyric acid (PBA) and salubrinal were used to inhibit ER stress and eIF2α dephosphorylation, respectively. We analyzed the eIF2α phosphorylation, ER stress, and hepatocyte necroptosis in mice and cells model. RESULTS: Tunicamycin or d-galactosamine significantly induced ER stress and necroptosis, as well as eIF2α phosphorylation, in mice and LO2 cells (p<0.05). ER stress aggravated tunicamycin-induced hepatocyte necroptosis in mice and LO2 cells (p<0.05). Elevated eIF2α phosphorylation significantly mitigated hepatocyte ER stress (p<0.05) and hepatocyte necroptosis in mice (34.37±3.39% vs 22.53±2.18%; p<0.05) and LO2 cells (1±0.11 vs 0.33±0.05; p<0.05). Interestingly, tumor necrosis factor receptor (TNFR) 1 protein levels were not completely synchronized with necroptosis. TNFR1 expression was reduced in d-galactosamine-treated mice (p<0.05) and cells incubated with tunicamycin for 12 and 24h (p<0.05). ER stress partially restored TNFR1 expression and increased necroptosis in tunicamycin-incubated cells (p<0.05). CONCLUSIONS: These results imply that ER stress can mediate hepatocyte necroptosis independent of TNFR1 signaling and elevated eIF2α phosphorylation can mitigate ER stress during acute liver injury.

Oxidative damage increased in presenilin1/presenilin2 conditional double knockout mice.

OBJECTIVE: This report aims to describe the oxidative damage profile in brain of presenilin1 and presenilin2 conditional double knockout mice (dKO) at both early and late age stages, and to discuss the correlation between oxidative stress and the Alzheimer-like phenotypes of dKO mice. METHODS: The protein level of Abeta(42) in dKO cortex and free 8-OHdG level in urine were measured by ELISA. Thiobarbituric acid method and spectrophotometric DNPH assay were used to determine the lipid peroxidation and protein oxidation in cortex, respectively. SOD and GSH-PX activities were assessed by SOD Assay Kit-WST and GSH-PX assay kit, separately. RESULTS: Significant decrease of Abeta(42) was verified in dKO cortex at 6 months as compared to control mice. Although lipid peroxidation (assessed by MDA) was increased only in dKO cortex at 3 months and protein oxidation (assessed by carbonyl groups) was basically unchanged in dKO cortex, ELISA analysis revealed that free 8-OHdG, which was an indicator of DNA lesion, was significantly decreased in urine of dKO mice from 3 months to 12 months. Activities of SOD and GSH-PX in dKO and control cortices showed no statistical difference except a significant increase of GSH-PX activity in dKO mice at 9 months. CONCLUSION: Oxidative damage, especially DNA lesion, was correlated with the neurodegenerative symptoms that appeared in dKO mice without the deposition of Abeta(42). Triggers of oxidative damage could be the inflammatory mediators released by activated microglia and astrocytes.