Search for: All records

Deep learning algorithms have been successfully adopted to extract meaningful information from digital images, yet many of them have been untapped in the semantic image segmentation of histopathology images. In this paper, we propose a deep convolutional neural network model that strengthens Atrous separable convolutions with a high rate within spatial pyramid pooling for histopathology image segmentation. A well-known model called DeepLabV3Plus was used for the encoder and decoder process. ResNet50 was adopted for the encoder block of the model which provides us the advantage of attenuating the problem of the increased depth of the network by using skip connections. Three Atrous separable convolutions with higher rates were added to the existing Atrous separable convolutions. We conducted a performance evaluation on three tissue types: tumor, tumor-infiltrating lymphocytes, and stroma for comparing the proposed model with the eight state-of-the-art deep learning models: DeepLabV3, DeepLabV3Plus, LinkNet, MANet, PAN, PSPnet, UNet, and UNet++. The performance results show that the proposed model outperforms the eight models on mIOU (0.8058/0.7792) and FSCR (0.8525/0.8328) for both tumor and tumor-infiltrating lymphocytes. 

Unraveling the mechanisms underlying the maintenance of species diversity is a central pursuit in ecology. It has been hypothesized that ectomycorrhizal (EcM) in contrast to arbuscular mycorrhizal fungi can reduce tree species diversity in local communities, which remains to be tested at the global scale. To address this gap, we analyzed global forest inventory data and revealed that the relationship between tree species richness and EcM tree proportion varied along environmental gradients. Specifically, the relationship is more negative at low latitudes and in moist conditions but is unimodal at high latitudes and in arid conditions. The negative association of EcM tree proportion on species diversity at low latitudes and in humid conditions is likely due to more negative plant-soil microbial interactions in these regions. These findings extend our knowledge on the mechanisms shaping global patterns in plant species diversity from a belowground view. 

Abstract Organic semiconductors are usually polycyclic aromatic hydrocarbons and their analogs containing heteroatom substitution. Bioinspired materials chemistry of organic electronics promises new charge transport mechanism and specific molecular recognition with biomolecules. We discover organic semiconductors from deoxyribonucleic acid topoisomerase inhibitors, featuring conjugated backbone decorated with hydrogen-bonding moieties distinct from common organic semiconductors. Using ellipticine as a model compound, we find that hydrogen bonds not only guide polymorph assembly, but are also critical to forming efficient charge transport pathways along π−conjugated planes when at a low dihedral angle by shortening the end-to-end distance of adjacent π planes. In the π−π stacking and hydrogen-bonding directions, the intrinsic, short-range hole mobilities reach as high as 6.5 cm²V⁻¹s⁻¹and 4.2 cm²V⁻¹s⁻¹measured by microwave conductivity, and the long-range apparent hole mobilities are up to 1.3 × 10^–3cm²V⁻¹s⁻¹and 0.4 × 10^–3cm²V⁻¹s⁻¹measured in field-effect transistors. We further demonstrate printed transistor devices and chemical sensors as potential applications.