Gastroesophageal reflux disease (GERD) is a common disease with high prevalence, and its endoscopic severity can be evaluated using the Los Angeles classification (LA grade). This paper proposes a deep learning model (i.e., GERD-VGGNet) that employs convolutional neural networks for automatic classification and interpretation of routine GERD LA grade. The proposed model employs a data augmentation technique, a two-stage no-freezing fine-tuning policy, and an early stopping criterion. As a result, the proposed model exhibits high generalizability. A dataset of images from 464 patients was used for model training and validation. An additional 32 patients served as a test set to evaluate the accuracy of both the model and our trainees. Experimental results demonstrate that the best model for the development set exhibited an overall accuracy of 99.2% (grade A–B), 100% (grade C–D), and 100% (normal group) using narrow-band image (NBI) endoscopy. On the test set, the proposed model resulted in an accuracy of 87.9%, which was significantly higher than the results of the trainees (75.0% and 65.6%). The proposed GERD-VGGNet model can assist automatic classification of GERD in conventional and NBI environments and thereby increase the accuracy of interpretation of the results by inexperienced endoscopists.

Diabetic neuropathy (DN) is a type of sensory nerve damage that can occur in patients with diabetes. Although the understanding of pathophysiology is incomplete, DN is often associated with structural and functional alterations of the affected neurons. Among all possible causes of nerve damage, Schwann cells (SCs) are thought to play a key role in repairing peripheral nerve injury, suggesting that functional deficits occurring in SCs may potentially exhibit their pathogenic roles in DN. Therefore, elucidating the mechanisms that underlie this pathology can be used to develop novel therapeutic targets. In this regard, glucagon-like peptide-1 receptor agonists (GLP-1 RAs) have recently attracted great attention in ameliorating SCs’ dysfunction. However, the detailed mechanisms remain uncertain. In the present study, we investigated how GLP-1 RA Liraglutide protects against RSC96 SCs dysfunction through a diabetic condition mimicked by high glucose and high free fatty acid (FFA). Our results showed that high glucose and high FFAs reduced the viability of RSC96 SCs by up to 51%, whereas Liraglutide reduced oxidative stress by upregulating antioxidant enzymes, and thus protected cells from apoptosis. Liraglutide also inhibited NFκB-mediated inflammation, inducing SCs to switch from pro-inflammatory cytokine production to anti-inflammatory cytokine production. Moreover, Liraglutide upregulated the production of neurotrophic factors and myelination-related proteins, and these protective effects appear to be synergistically linked to insulin signaling. Taken together, our findings demonstrate that Liraglutide ameliorates diabetes-related SC dysfunction through the above-mentioned mechanisms, and suggest that modulating GLP-1 signaling in SCs may be a promising strategy against DN.

Obesity is strongly associated with insulin sensitivity in type 2 diabetes (T2D), mainly because free fatty acids (FFAs) are released from excess fat tissue. Long-term exposure to high levels of FFAs and glucose leads to glucolipotoxicity, causing damage to pancreatic β-cells, thus accelerating the progression of T2D. Therefore, the prevention of β-cell dysfunction and apoptosis is essential to prevent the development of T2D. Unfortunately, there are currently no specific clinical strategies for protecting β-cells, highlighting the need for effective therapies or preventive approaches to improve the survival of β-cells in T2D. Interestingly, recent studies have shown that the monoclonal antibody denosumab (DMB), used in osteoporosis, displays a positive effect on blood glucose regulation in patients with T2D. DMB acts as an osteoprotegerin (OPG) by inhibiting the receptor activator of the NF-κB ligand (RANKL), preventing the maturation and function of osteoclasts. However, the exact mechanism by which the RANK/RANKL signal affects glucose homeostasis has not been fully explained. The present study used human 1.4 × 10⁷ β-cells to simulate the T2D metabolic condition of high glucose and free fatty acids (FFAs), and it investigated the ability of DMB to protect β-cells from glucolipotoxicity. Our results show that DMB effectively attenuated the cell dysfunction and apoptosis caused by high glucose and FFAs in β-cells. This may be caused by blocking the RANK/RANKL pathway that reduced mammalian sterile 20-like kinase 1 (MST1) activation and indirectly increased pancreatic and duodenal homeobox 1 (PDX-1) expression. Furthermore, the increase in inflammatory cytokines and ROS caused by the RANK/RANKL signal also played an important role in glucolipotoxicity-induced cytotoxicity, and DMB can also protect β-cells by reducing the mechanisms mentioned above. These findings provide detailed molecular mechanisms for the future development of DMB as a potential protective agent of β-cells.