Search for: All records

Supervised deep-learning models have enabled super-resolution imaging in several microscopic imaging modalities, increasing the spatial lateral bandwidth of the original input images beyond the diffraction limit. Despite their success, their practical application poses several challenges in terms of the amount of training data and its quality, requiring the experimental acquisition of large, paired databases to generate an accurate generalized model whose performance remains invariant to unseen data. Cycle-consistent generative adversarial networks (cycleGANs) are unsupervised models for image-to-image translation tasks that are trained on unpaired datasets. This paper introduces a cycleGAN framework specifically designed to increase the lateral resolution limit in confocal microscopy by training a cycleGAN model using low- and high-resolution unpaired confocal images of human glioblastoma cells. Training and testing performances of the cycleGAN model have been assessed by measuring specific metrics such as background standard deviation, peak-to-noise ratio, and a customized frequency content measure. Our cycleGAN model has been evaluated in terms of image fidelity and resolution improvement using a paired dataset, showing superior performance than other reported methods. This work highlights the efficacy and promise of cycleGAN models in tackling super-resolution microscopic imaging without paired training, paving the path for turning home-built low-resolution microscopic systems into low-cost super-resolution instruments by means of unsupervised deep learning. 

One of the major drawbacks of confocal microscopy is its limited spatial resolution. This work assesses the performance of an unpaired learning-based model to provide confocal images with improved resolution. 

Headwater streams are reliant on riparian tree leaf litterfall to fuel brown food webs. Terrestrial agents like herbivores and contaminants can alter plant growth, litter production, litter quality, and the timing of litterfall into streams, influencing aspects of the brown food web. At Mount St. Helens (USA), early successional streams are developing willow (Salix sitchensis) riparian zones. The willows are attacked by stem-boring herbivores, altering litter quality and the timing of litterfall. Within a established experimental plots, willows (male and female plants) were protected from herbivores using insecticides and provided with experimental additions of nitrogen. This enabled us to test the interacting influences of herbivores, nitrogen deposition, and willow sex on leaf litter quality, aquatic litter decomposition, and microbial and invertebrate detritivores. We found weak litter quality effects (higher N and lower C:N) for the herbivore treatment, but no effect of nitrogen deposition. Although litter decomposition rates were not strongly affected by litter treatments, detritivore communities were altered by all treatments. Nitrogen deposition resulted in decreased bacterial richness and decreased fungal diversity in-stream. Aquatic macroinvertebrate communities were influenced by the interacting effects of herbivory and nitrogen addition, with abundances highest in herbivore litter with the greatest N addition. Shredders showed the highest abundance in male, herbivore-attacked litter. The establishment of riparian willows along early successional streams and their interacting effects with herbivores and nitrogen deposition may be influencing detritivore community assembly at Mount St. Helens. More broadly, global changes like increased wet and dry N deposition and expanded ranges of key herbivores might influence tree litter decomposition in many ecosystems. 

Abstract Determining how streams develop naturally, particularly the ecological role of newly developed riparian canopy cover, is essential to understanding the factors that structure new stream communities and provides valuable information for restoring highly disturbed ecosystems. However, attempts to understand primary succession in riverine ecosystems have been hindered by a lack of data owing to the infrequent formation of new rivers on the landscape. In the present study, we used five streams formed following the 1980 eruption of Mount St. Helens (WA, USA) to examine the influence of canopy cover development on algal and benthic macroinvertebrate assemblages, biomass, and organic matter processing. Newly established closed canopy reaches had less available light, but no significant differences in algal biomass or macroinvertebrate assemblages compared to open canopy reaches. Instead, algal and macroinvertebrate communities were structured mainly by hydrologic differences among watersheds. In contrast, organic matter processing rates were sensitive to canopy cover development, and rates were faster under closed canopies, especially in late summer or after terrestrial preconditioning. After 40 years of stream and riparian primary successional development, canopy cover strongly influences ecosystem function, but aquatic organism assembly was more influenced by physio-chemical and hydrologic variation. Our findings provide insight into the development of in-stream assemblages and ecosystem functions, which is also relevant to efforts to address major disturbances to stream channels, such as volcanic eruptions, floods, forest fires, and clear-cut logging.