Deep learning framework for rapid and accurate respiratory COVID-19 prediction using chest X-ray images.

COVID-19 is a contagious disease that affects the human respiratory system. Infected individuals may develop serious illnesses, and complications may result in death. Using medical images to detect COVID-19 from essentially identical thoracic anomalies is challenging because it is time-consuming, laborious, and prone to human error. This study proposes an end-to-end deep-learning framework based on deep feature concatenation and a Multi-head Self-attention network. Feature concatenation involves fine-tuning the pre-trained backbone models of DenseNet, VGG-16, and InceptionV3, which are trained on a large-scale ImageNet, whereas a Multi-head Self-attention network is adopted for performance gain. End-to-end training and evaluation procedures are conducted using the COVID-19_Radiography_Dataset for binary and multi-classification scenarios. The proposed model achieved overall accuracies (96.33% and 98.67%) and F1_scores (92.68% and 98.67%) for multi and binary classification scenarios, respectively. In addition, this study highlights the difference in accuracy (98.0% vs. 96.33%) and F_1 score (97.34% vs. 95.10%) when compared with feature concatenation against the highest individual model performance. Furthermore, a virtual representation of the saliency maps of the employed attention mechanism focusing on the abnormal regions is presented using explainable artificial intelligence (XAI) technology. The proposed framework provided better COVID-19 prediction results outperforming other recent deep learning models using the same dataset.

Exosomes/microvesicles target SARS-CoV-2 via innate and RNA-induced immunity with PIWI-piRNA system.

Murine neural stem cells (NSCs) were recently shown to release piRNA-containing exosomes/microvesicles (Ex/Mv) for exerting antiviral immunity, but it remains unknown if these Ex/Mv could target SARS-CoV-2 and whether the PIWI-piRNA system is important for these antiviral actions. Here, using in vitro infection models, we show that hypothalamic NSCs (htNSCs) Ex/Mv provided an innate immunity protection against SARS-CoV-2. Importantly, enhanced antiviral actions were achieved by using induced Ex/Mv that were derived from induced htNSCs through twice being exposed to several RNA fragments of SARS-CoV-2 genome, a process that was designed not to involve protein translation of these RNA fragments. The increased antiviral effects of these induced Ex/Mv were associated with increased expression of piRNA species some of which could predictably target SARS-CoV-2 genome. Knockout of piRNA-interacting protein PIWIL2 in htNSCs led to reductions in both innate and induced antiviral effects of Ex/Mv in targeting SARS-CoV-2. Taken together, this study demonstrates a case suggesting Ex/Mv from certain cell types have innate and adaptive immunity against SARS-CoV-2, and the PIWI-piRNA system is important for these antiviral actions.

Innate and Adaptive Immunity of Murine Neural Stem Cell-Derived piRNA Exosomes/Microvesicles against Pseudotyped SARS-CoV-2 and HIV-Based Lentivirus.

By testing pseudotyped SARS-CoV-2 and HIV-based lentivirus, this study reports that exosomes/microvesicles (Ex/Mv) isolated from murine hypothalamic neural stem/progenitor cells (htNSC) or subtype htNSCPGHM as well as hippocampal NSC have innate immunity-like actions against these RNA viruses. These extracellular vesicles also have a cell-free innate antiviral action by attacking and degrading viruses. We further generated the induced versions of Ex/Mv through prior viral exposure to NSCs and found that these induced Ex/Mv were stronger than basal Ex/Mv in reducing the infection of these viruses, suggesting the involvement of an adaptive immunity-like antiviral function. These NSC Ex/Mv were found to be characterized by producing large libraries of P element-induced wimpy testis (PIWI)-interacting RNAs (piRNAs) against genomes of various viruses, and some of these piRNAs were enriched during the adaptive immunity-like reaction, possibly contributing to the antiviral effects of these Ex/Mv. In conclusion, NSC Ex/Mv have antiviral immunity and could potentially be developed to combat against various viruses.