Search

Recently, convolutional neural networks (CNNs) and self-attention mechanisms have been widely applied in plant disease identification tasks, yielding significant successes. Currently, the majority of research models for tomato leaf disease recognition rely solely on traditional convolutional models or Transformer architectures and fail to capture both local and global features simultaneously. This limitation may result in biases in the model's focus, consequently impacting the accuracy of disease recognition. Consequently, models capable of extracting local features while attending to global information have emerged as a novel research direction. To address these challenges, we propose an Eff-Swin model that integrates the enhanced features of the EfficientNetV2 and Swin Transformer networks, aiming to harness the local feature extraction capability of CNNs and the global modeling ability of Transformers. Comparative experiments demonstrate that the enhanced model has achieved a further increase in training accuracy, reaching an accuracy rate of 99.70% on the tomato leaf disease dataset, which is 0.49~3.68% higher than that of individual network models and 0.8~1.15% higher than that of existing state-of-the-art combined approaches. The results show that integrating attention mechanisms into convolutional models can significantly enhance the accuracy of tomato leaf disease recognition while also offering the great potential of the Eff-Swin backbone with self-attention in plant disease identification. [ABSTRACT FROM AUTHOR]

The regulation of human pluripotent stem cell (hPSC) behaviors has been mainly studied through exploration of biochemical factors. However, the current directed differentiation protocols for hPSCs that completely rely on biochemical factors remain suboptimal. It has recently become evident that coexisting biophysical signals in the stem cell microenvironment, including nanotopographic cues, can provide potent regulatory signals to mediate adult stem cell behaviors, including self-renewal and differentiation. Herein, we utilized a recently developed, large-scale nanofabrication technique based on reactive-ion etching (RIE) to generate random nanoscale structures on glass surfaces with high precision and reproducibility. We report here that hPSCs are sensitive to nanotopographic cues and such nanotopographic sensitivity can be leveraged for improving directed neuronal differentiation of hPSCs. We demonstrate early neuroepithelial conversion and motor neuron (MN) progenitor differentiation of hPSCs can be promoted using nanoengineered topographic substrates. We further explore how hPSCs sense the substrate nanotopography and relay this biophysical signal through a regulatory signaling network involving cell adhesion, the actomyosin cytoskeleton, and Hippo/YAP signaling to mediate the neuroepithelial induction of hPSCs. Our study provides an efficient method for large-scale production of MNs from hPSCs, useful for regenerative medicine and cell-based therapies.

Our understanding of the intrinsic mechanosensitive properties of human pluripotent stem cells (hPSCs), in particular the effects that the physical microenvironment has on their differentiation, remains elusive. Here, we show that neural induction and caudalization of hPSCs can be accelerated by using a synthetic microengineered substrate system consisting of poly(dimethylsiloxane) micropost arrays (PMAs) with tunable mechanical rigidities. The purity and yield of functional motor neurons derived from hPSCs within 23 days of culture using soft PMAs were improved more than fourfold and tenfold, respectively, compared with coverslips or rigid PMAs. Mechanistic studies revealed a multi-targeted mechanotransductive process involving Smad phosphorylation and nucleocytoplasmic shuttling, regulated by rigidity-dependent Hippo/YAP activities and actomyosin cytoskeleton integrity and contractility. Our findings suggest that substrate rigidity is an important biophysical cue influencing neural induction and subtype specification, and that microengineered substrates can thus serve as a promising platform for large-scale culture of hPSCs.