ECG arrhythmia detection in an inter-patient setting using Fourier decomposition and machine learning.

ECG beat classification or arrhythmia detection through artificial intelligence (AI) is an active topic of research. It is vital to recognize and detect the type of arrhythmia for monitoring cardiac abnormalities. The AI-based ECG beat classification algorithms proposed in the literature suffer from two main drawbacks. Firstly, some of the works have not considered any unseen test data to validate the performance of their algorithms. Secondly, the accuracy of detecting superventricular ectopic beats (SVEB) needs to be improved. In this work, we address these issues by considering an inter-patient paradigm where the test dataset is collected from a different set of subjects than the training data. Also, the proposed methodology detects SVEB with an F1 score of 89.35%, which is better than existing algorithms. We have used the Fourier decomposition method (FDM) for multi-scale analysis of ECG signals and extracted time-domain and statistical features from the narrow-band signal components obtained using FDM. Feature selection techniques, including the Kruskal-Wallis test and minimum redundancy maximum relevance (mRMR) have been used to select only the relevant features and rank these features to remove any redundancy. Since the dataset used is highly imbalanced, Mathew's correlation coefficient (MCC) has also been used to analyze the performance of the proposed method. Support vector machine classifier with linear kernel achieves an overall 98.03% accuracy and 91.84% MCC for the MIT-BIH arrhythmia dataset.

A two-stage transformer based network for motor imagery classification.

Brain-computer interfaces (BCIs) are used to understand brain functioning and develop therapies for neurological and neurodegenerative disorders. Therefore, BCIs are crucial in rehabilitating motor dysfunction and advancing motor imagery applications. For motor imagery, electroencephalogram (EEG) signals are used to classify the subject's intention of moving a body part without actually moving it. This paper presents a two-stage transformer-based architecture that employs handcrafted features and deep learning techniques to enhance the classification performance on benchmarked EEG signals. Stage-1 is built on parallel convolution based EEGNet, multi-head attention, and separable temporal convolution networks for spatiotemporal feature extraction. Further, for enhanced classification, in stage-2, additional features and embeddings extracted from stage-1 are used to train TabNet. In addition, a novel channel cluster swapping data augmentation technique is also developed to handle the issue of limited samples for training deep learning architectures. The developed two-stage architecture offered an average classification accuracy of 88.5 % and 88.3 % on the BCI Competition IV-2a and IV-2b datasets, respectively, which is approximately 3.0 % superior over similar recent reported works.

Rv0495c regulates redox homeostasis in Mycobacterium tuberculosis.

Mycobacterium tuberculosis (Mtb) has evolved sophisticated surveillance mechanisms to neutralize the ROS-induces toxicity which otherwise would degrade a variety of biological molecules including proteins, nucleic acids and lipids. In the present study, we find that Mtb lacking the Rv0495c gene (ΔRv0495c) is presented with a highly oxidized cytosolic environment. The superoxide-induced lipid peroxidation resulted in altered colony morphology and loss of membrane integrity in ΔRv0495c. As a consequence, ΔRv0495c demonstrated enhanced susceptibility when exposed to various host-induced stress conditions. Further, as expected, we observed a mutant-specific increase in the abundance of transcripts that encode proteins involved in antioxidant defence. Surprisingly, despite showing a growth defect phenotype in macrophages, the absence of the Rv0495c enhanced the pathogenicity and augmented the ability of the Mtb to grow inside the host. Additionally, our study revealed that Rv0495c-mediated immunomodulation by the pathogen helps create a favorable niche for long-term survival of Mtb inside the host. In summary, the current study underscores the fact that the truce in the war between the host and the pathogen favours long-term disease persistence in tuberculosis. We believe targeting Rv0495c could potentially be explored as a strategy to potentiate the current anti-TB regimen.

Fetal liver CD34+ contain human immune and endothelial progenitors and mediate solid tumor rejection in NOG mice.

BACKGROUND: Transplantation of CD34+ hematopoietic stem and progenitor cells (HSPC) into immunodeficient mice is an established method to generate humanized mice harbouring a human immune system. Different sources and methods for CD34+ isolation have been employed by various research groups, resulting in customized models that are difficult to compare. A more detailed characterization of CD34+ isolates is needed for a better understanding of engraftable hematopoietic and potentially non-hematopoietic cells. Here we have performed a direct comparison of CD34+ isolated from cord blood (CB-CD34+) or fetal liver (FL-CD34+ and FL-CD34+CD14-) and their engraftment into immunocompromised NOD/Shi-scid Il2rgnull (NOG) mice. METHODS: NOG mice were transplanted with either CB-CD34+, FL-CD34+ or FL-CD34+CD14- to generate CB-NOG, FL-NOG and FL-CD14--NOG, respectively. After 15-20 weeks, the mice were sacrificed and human immune cell reconstitution was assessed in blood and several organs. Liver sections were pathologically assessed upon Haematoxylin and Eosin staining. To assess the capability of allogenic tumor rejection in CB- vs. FL-reconstituted mice, animals were subcutaneously engrafted with an HLA-mismatched melanoma cell line. Tumor growth was assessed by calliper measurements and a Luminex-based assay was used to compare the cytokine/chemokine profiles. RESULTS: We show that CB-CD34+ are a uniform population of HSPC that reconstitute NOG mice more rapidly than FL-CD34+ due to faster B cell development. However, upon long-term engraftment, FL-NOG display increased numbers of neutrophils, dendritic cells and macrophages in multiple tissues. In addition to HSPC, FL-CD34+ isolates contain non-hematopoietic CD14+ endothelial cells that enhance the engraftment of the human immune system in FL-NOG mice. We demonstrate that these CD14+CD34+ cells are capable of reconstituting Factor VIII-producing liver sinusoidal endothelial cells (LSEC) in FL-NOG. However, CD14+CD34+ also contribute to hepatic sinusoidal dilatation and immune cell infiltration, which may culminate in a graft-versus-host disease (GVHD) pathology upon long-term engraftment. Finally, using an HLA-A mismatched CDX melanoma model, we show that FL-NOG, but not CB-NOG, can mount a graft-versus-tumor (GVT) response resulting in tumor rejection. CONCLUSION: Our results highlight important phenotypical and functional differences between CB- and FL-NOG and reveal FL-NOG as a potential model to study hepatic sinusoidal dilatation and mechanisms of GVT.

Longitudinal single cell atlas identifies complex temporal relationship between type I interferon response and COVID-19 severity.

Due to the paucity of longitudinal molecular studies of COVID-19, particularly those covering the early stages of infection (Days 1-8 symptom onset), our understanding of host response over the disease course is limited. We perform longitudinal single cell RNA-seq on 286 blood samples from 108 age- and sex-matched COVID-19 patients, including 73 with early samples. We examine discrete cell subtypes and continuous cell states longitudinally, and we identify upregulation of type I IFN-stimulated genes (ISGs) as the predominant early signature of subsequent worsening of symptoms, which we validate in an independent cohort and corroborate by plasma markers. However, ISG expression is dynamic in progressors, spiking early and then rapidly receding to the level of severity-matched non-progressors. In contrast, cross-sectional analysis shows that ISG expression is deficient and IFN suppressors such as SOCS3 are upregulated in severe and critical COVID-19. We validate the latter in four independent cohorts, and SOCS3 inhibition reduces SARS-CoV-2 replication in vitro. In summary, we identify complexity in type I IFN response to COVID-19, as well as a potential avenue for host-directed therapy.