Purpose . To evaluate the peripapillary vessel density (pVD) and the peripapillary nerve fiber layer (pRNFL) thickness in patients with chiasmal compression caused by sellar region mass using optical coherence tomography angiography (OCTA). Methods . This is an observational, cross‐sectional study of 31 patients (31 eyes) with chiasmal compression caused by sellar region mass and 34 healthy controls (34 eyes). Automated perimetry and OCTA were performed. The pVD and pRNFL thickness were compared between the two groups. The impact of tumor diameter, duration of symptoms, and cavernous sinus (CS) invasion on visual dysfunction, pVD, and pRNFL thickness was also analyzed. Furthermore, we divided the patients into two subgroups according to whether there was an absolute defect in the central visual field and evaluated their pVD and pRNFL thickness, respectively. Results . Compared to the healthy control group, there was a statistically significant decrease in pVD and pRNFL thickness in patients with chiasmal compression ( p < 0.05), especially in patients with substantial absolute defects in the central visual field. Tumor diameter, duration of symptoms, and CS invasion did not appear to be associated with pVD and pRNFL thickness. There was a significant positive correlation between the pVD and pRNFL thickness in patients with chiasmal compression ( p < 0.001). Conclusion . pVD and pRNFL thickness are significantly decreased in patients with chiasmal compression revealed by OCTA, especially in patients with more severe visual field defects. A significant correlation between pVD and pRNFL thickness was demonstrated, which provides a clue for the study of the mechanism of changes in retinal perfusion in compressive optic neuropathy. It requires considerable attention that OCTA may play an important role in disease monitoring of sellar region mass. Hence, further studies are needed to verify whether OCTA is helpful to predict the prognosis of visual function after decompression surgery.

Diabetic retinopathy (DR) is the largest cause of permanent vision loss in the working-age population, making automated grading critical for timely therapeutic intervention. While recent deep learning algorithms have improved feature discrimination, modern state-of-the-art systems have two fundamental drawbacks. First, most models rely on standard Convolutional Neural Networks, which struggle to capture long-range relationships and lack semantic reasoning, resulting in visual findings that do not correlate with clinical knowledge. Second, present approaches often consider grading as a nominal classification or a pure ordinal regression task, failing to strike a compromise between high classification accuracy and severity-consistent predictions (Quadratic Weighted Kappa). To address these challenges, we propose Dual-SwinOrd, a novel framework that integrates a hierarchical Vision Transformer with a semantically guided dual-head mechanism. Specifically, we use a Swin Transformer backbone to extract hierarchical features, effectively capturing global retinal structures. To handle diverse lesion scales, we incorporate a Progressive Lesion-aware Kernel Attention (PLKA) module and a Semantic Prior Modulation (SPM) module guided by PubMedCLIP, bridging the gap between visual features and medical linguistic priors. In addition, we propose a Dual-Head learning strategy that decouples the optimization objective into two parallel streams: a Classification Head to maximize diagnostic accuracy and an Ordinal Regression Head (DPE) to enforce rank-consistency. This design effectively mitigates the trade-off between precision and ordinality. Extensive experiments on the APTOS 2019 and DDR datasets demonstrate that Dual-SwinOrd achieves state-of-the-art performance, yielding an Accuracy of 87.98% and a Quadratic Weighted Kappa (QWK) of 0.9370 on the APTOS 2019 dataset, as well as an Accuracy of 86.54% and a QWK of 0.9040 on the DDR dataset.