Yijie Lin

Purpose: To compare perfused peripapillary capillary density in primary open-angle glaucoma (POAG), normal-tension glaucoma (NTG), and normal patients using optical coherence tomography angiography (OCT-A). Methods: A retrospective review of POAG, NTG, and normal patients imaged with OCT-A was performed. En face OCT angiograms identifying peripapillary vessels were obtained using a spectral-domain OCT system (Avanti RTVue-XR). A custom image analysis approach identified perfused peripapillary capillaries, quantified perfused capillary density (PCD), and generated color-coded PCD maps for 3.5- and 4.5-mm-diameter scans. We compared PCD values, PCD maps, standard automated perimetry (Humphrey visual field [HVF]) parameters, and OCT retinal nerve fiber layer (RNFL) thickness analyses across all groups. Results: Forty POAG, 26 NTG, and 26 normal patients were included. Annular PCD in POAG (34.24 ± 6.76%) and NTG (37.75 ± 3.52%) patients was significantly decreased compared to normal patients (42.99 ± 1.81%) in 4.5-mm scans (P < 0.01 and P < 0.01, respectively). Similar trends and statistical significances were seen in 3.5-mm scans. Linear regression analysis resulted in moderate correlations between annular PCD values and other glaucomatous parameters. Pearson coefficients comparing annular PCD from 4.5-mm scans in POAG and NTG groups to HVF mean deviation, HVF pattern standard deviation, and average RNFL thickness all showed statistical significance (P < 0.05). Color maps showed that POAG and NTG patients had a reduction of perfused capillaries that progressed in size when comparing early, moderate, and severe glaucoma groups. Conclusions: Optical coherence tomography angiography can uniquely identify changes in peripapillary PCD in glaucoma patients. Optical coherence tomography angiography may offer insights into the pathophysiology of glaucomatous damage and risk factors for disease progression.

CuO particles can be used as an agrochemical with slow and controlled release of micronutrients but, their size-specific effects have not been well elucidated. In this study, soil-grown lettuce was exposed to CuO nanoparticles (nCuO) and CuO microparticles (μCuO) at 0, 200, and 400 mg Cu/kg for 60 days. At harvest, nCuO released 82.2% and 117.7% more Cu ions, as well as 48.7% and 75.3% more DTPA-extractable Cu than μCuO at 200 and 400 mg/kg, respectively. CuO treatments induced higher copper accumulated in lettuce roots but did not change shoot’s Cu content or cause oxidative stress. nCuO treatments increased shoot’s biomass by 16.3–19.1%, while μCuO showed no improvement to the biomass. The beneficial effects on lettuce by nCuO may be attributed to its modification in plant photosynthesis system with increased transpiration rate and higher stomatal conductance. Results indicate that nCuO can be a promising agrochemical fertilizer to increase crop production.

In recent, some robust contrastive multi-view clustering (MvC) methods have been proposed, which construct data pairs from neighborhoods to alleviate the false negative issue, i.e., some intra-cluster samples are wrongly treated as negative pairs. Although promising performance has been achieved by these methods, the false negative issue is still far from addressed and the false positive issue emerges because all in- and out-of-neighborhood samples are simply treated as positive and negative, respectively. To address the issues, we propose a novel robust method, dubbed decoupled contrastive multi-view clustering with high-order random walks (DIVIDE). In brief, DIVIDE leverages random walks to progressively identify data pairs in a global instead of local manner. As a result, DIVIDE could identify in-neighborhood negatives and out-of-neighborhood positives. Moreover, DIVIDE embraces a novel MvC architecture to perform inter- and intra-view contrastive learning in different embedding spaces, thus boosting clustering performance and embracing the robustness against missing views. To verify the efficacy of DIVIDE, we carry out extensive experiments on four benchmark datasets comparing with nine state-of-the-art MvC methods in both complete and incomplete MvC settings. The code is released on https://github.com/XLearning-SCU/2024-AAAI-DIVIDE.