Fitbeat: COVID-19 estimation based on wristband heart rate using a contrastive convolutional auto-encoder.
Ontology highlight
ABSTRACT:
SUBMITTER: Liu S
PROVIDER: S-EPMC8547790 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:Steganalysis is the process of analyzing and predicting the presence of hidden information in images. Steganalysis would be most useful to predict whether the received images contain useful information. However, it is more difficult to predict the hidden information in images which is computationally difficult. In the existing research method, this is resolved by introducing the deep learning approach which attempts to perform steganalysis tasks in effectively. However, this research method does not concentrate the noises present in the images. It might increase the computational overhead where the error cost adjustment would require more iteration. This is resolved in the proposed research technique by introducing the novel research method called Non-Gaussian Noise Aware Auto Encoder Convolutional Neural Network (NGN-AEDNN). Classification technique provides a more flexible way for steganalysis where the multiple features present in the environment would lead to an inaccurate prediction rate. Here, learning accuracy is improved by introducing noise removal techniques before performing a learning task. Non-Gaussian Noise Removal technique is utilized to remove the noises before learning. Also, Gaussian noise removal is applied at every iteration of the neural network to adjust the error rate without the involvement of noisy features. This proposed work can ensure efficient steganalysis by accurate learning task. Matlab has been employed to implement the method by performing simulations from which it is proved that the proposed research technique NGN-AEDNN can ensure the efficient steganalysis outcome with the reduced computational overhead when compared with the existing methods.
| S-EPMC7959617 | biostudies-literature
Project description:Recent advances in spatial transcriptomics (ST) have enabled comprehensive profiling of gene expression with spatial information in the context of the tissue microenvironment. However, with the improvements in the resolution and scale of ST data, deciphering spatial domains precisely while ensuring efficiency and scalability is still challenging. Here, we develop SGCAST, an efficient auto-encoder framework to identify spatial domains. SGCAST adopts a symmetric graph convolutional auto-encoder to learn aggregated latent embeddings via integrating the gene expression similarity and the proximity of the spatial spots. This framework in SGCAST enables a mini-batch training strategy, which makes SGCAST memory-efficient and scalable to high-resolution spatial transcriptomic data with a large number of spots. SGCAST improves the overall accuracy of spatial domain identification on benchmarking data. We also validated the performance of SGCAST on ST datasets at various scales across multiple platforms. Our study illustrates the superior capacity of SGCAST on analyzing spatial transcriptomic data.
| S-EPMC10782917 | biostudies-literature
Project description:This work presents the application of an Encoder-Decoder convolutional neural network (ED-CNN) model to automatically segment COVID-19 computerised tomography (CT) data. By doing so we are producing an alternative model to current literature, which is easy to follow and reproduce, making it more accessible for real-world applications as little training would be required to use this. Our simple approach achieves results comparable to those of previously published studies, which use more complex deep-learning networks. We demonstrate a high-quality automated segmentation prediction of thoracic CT scans that correctly delineates the infected regions of the lungs. This segmentation automation can be used as a tool to speed up the contouring process, either to check manual contouring in place of a peer checking, when not possible or to give a rapid indication of infection to be referred for further treatment, thus saving time and resources. In contrast, manual contouring is a time-consuming process in which a professional would contour each patient one by one to be later checked by another professional. The proposed model uses approximately 49 k parameters while others average over 1,000 times more parameters. As our approach relies on a very compact model, shorter training times are observed, which make it possible to easily retrain the model using other data and potentially afford "personalised medicine" workflows. The model achieves similarity scores of Specificity (Sp) = 0.996 ± 0.001, Accuracy (Acc) = 0.994 ± 0.002 and Mean absolute error (MAE) = 0.0075 ± 0.0005.
| S-EPMC11323195 | biostudies-literature
Project description:This paper gives a solution to the blind parameter identification of a convolutional encoder. The problem can be addressed in the context of the noncooperative communications or adaptive coding and modulations (ACM) for cognitive radio networks. We consider an intelligent communication receiver which can blindly recognize the coding parameters of the received data stream. The only knowledge is that the stream is encoded using binary convolutional codes, while the coding parameters are unknown. Some previous literatures have significant contributions for the recognition of convolutional encoder parameters in hard-decision situations. However, soft-decision systems are applied more and more as the improvement of signal processing techniques. In this paper we propose a method to utilize the soft information to improve the recognition performances in soft-decision communication systems. Besides, we propose a new recognition method based on correlation attack to meet low signal-to-noise ratio situations. Finally we give the simulation results to show the efficiency of the proposed methods.
| S-EPMC4055125 | biostudies-other
Project description:PurposeCoronavirus disease 2019 (COVID-19) is caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), which primarily infects the lower airways and binds to angiotensin-converting enzyme 2 (ACE2) on alveolar epithelial cells. ACE2 is widely expressed not only in the lungs but also in the cardiovascular system. Therefore, SARS-CoV-2 can also damage the myocardium. This report aimed to highlight decreased heart rate variability (HRV) and cardiac injury caused by SARS-CoV-2.Materials and methodsWe evaluated three COVID-19 patients who died. Patients' data were collected from electronic medical records. We collected patient's information, including baseline information, lab results, body temperature, heart rate (HR), clinical outcome and other related data. We calculated the HRV and the difference between the expected and actual heart rate changes as the body temperature increased.ResultsAs of March 14, 2020, 3 (2.2%) of 136 patients with COVID-19 in Tianjin died in the early stage of the COVID-19 epidemic. The immediate cause of death for Case 1, Case 2, and Case 3 was cardiogenic shock, cardiac arrest and cardiac arrest, respectively. The HRV were substantially decreased in the whole course of all three cases. The actual increases in heart rate were 5 beats/min, 13 beats/min, and 4 beats/min, respectively, less than expected as their temperature increased. Troponin I and Creatine Kinase MB isoenzyme (CK-MB) were substantially increased only in Case 3, for whom the diagnosis of virus-related cardiac injury could not be made until day 7. In all three cases, decreased in HRV and HR changes occurred earlier than increases in cardiac biomarkers (e.g., troponin I and CK-MB).ConclusionsIn conclusion, COVID-19 could affect HRV and counteract tachycardia in response to increases in body temperature. The decreases of HRV and HR changes happened earlier than the increases of myocardial markers (troponin I and CK-MB). It suggested the decreases of HRV and HR changes might help predict cardiac injury earlier than myocardial markers in COVID-19, thus its early identification might help improve patient prognosis.Supplementary informationThe online version contains supplementary material available at 10.1007/s44231-022-00024-1.
| S-EPMC9713139 | biostudies-literature
Project description:IntroductionStrict lockdown rules were imposed to the French population from 17 March to 11 May 2020, which may result in limited possibilities of physical activity, modified psychological and health states. This report is focused on HRV parameters kinetics before, during and after this lockdown period.Methods95 participants were included in this study (27 women, 68 men, 37 ± 11 years, 176 ± 8 cm, 71 ± 12 kg), who underwent regular orthostatic tests (a 5-minute supine followed by a 5-minute standing recording of heart rate (HR)) on a regular basis before (BSL), during (CFN) and after (RCV) the lockdown. HR, power in low- and high-frequency bands (LF, HF, respectively) and root mean square of the successive differences (RMSSD) were computed for each orthostatic test, and for each position. Subjective well-being was assessed on a 0-10 visual analogic scale (VAS). The participants were split in two groups, those who reported an improved well-being (WB+, increase >2 in VAS score) and those who did not (WB-) during CFN.ResultsOut of the 95 participants, 19 were classified WB+ and 76 WB-. There was an increase in HR and a decrease in RMSSD when measured supine in CFN and RCV, compared to BSL in WB-, whilst opposite results were found in WB+ (i.e. decrease in HR and increase in RMSSD in CFN and RCV; increase in LF and HF in RCV). When pooling data of the three phases, there were significant correlations between VAS and HR, RMSSD, HF, respectively, in the supine position; the higher the VAS score (i.e., subjective well-being), the higher the RMSSD and HF and the lower the HR. In standing position, HRV parameters were not modified during CFN but RMSSD was correlated to VAS.ConclusionOur results suggest that the strict COVID-19 lockdown likely had opposite effects on French population as 20% of participants improved parasympathetic activation (RMSSD, HF) and rated positively this period, whilst 80% showed altered responses and deteriorated well-being. The changes in HRV parameters during and after the lockdown period were in line with subjective well-being responses. The observed recordings may reflect a large variety of responses (anxiety, anticipatory stress, change on physical activity…) beyond the scope of the present study. However, these results confirmed the usefulness of HRV as a non-invasive means for monitoring well-being and health in this population.
| S-EPMC7660489 | biostudies-literature
Project description:Wearable monitoring devices are an innovative way to measure heart rate (HR) and heart rate variability (HRV), however, there is still debate about the validity of these wearables. This study aimed to validate the accuracy and predictive value of the Empatica E4 wristband against the VU University Ambulatory Monitoring System (VU-AMS) in a clinical population of traumatized adolescents in residential care. A sample of 345 recordings of both the Empatica E4 wristband and the VU-AMS was derived from a feasibility study that included fifteen participants. They wore both devices during two experimental testing and twelve intervention sessions. We used correlations, cross-correlations, Mann-Whitney tests, difference factors, Bland-Altman plots, and Limits of Agreement to evaluate differences in outcomes between devices. Significant correlations were found between Empatica E4 and VU-AMS recordings for HR, SDNN, RMSSD, and HF recordings. There was a significant difference between the devices for all parameters but HR, although effect sizes were small for SDNN, LF, and HF. For all parameters but RMSSD, testing outcomes of the two devices led to the same conclusions regarding significance. The Empatica E4 wristband provides a new opportunity to measure HRV in an unobtrusive way. Results of this study indicate the potential of the Empatica E4 as a practical and valid tool for research on HR and HRV under non-movement conditions. While more research needs to be conducted, this study could be considered as a first step to support the use of HRV recordings provided by wearables.
| S-EPMC7511462 | biostudies-literature
Project description:Recently, there have been great interests for computer-aided diagnosis of Alzheimer's disease (AD) and its prodromal stage, mild cognitive impairment (MCI). Unlike the previous methods that considered simple low-level features such as gray matter tissue volumes from MRI, and mean signal intensities from PET, in this paper, we propose a deep learning-based latent feature representation with a stacked auto-encoder (SAE). We believe that there exist latent non-linear complicated patterns inherent in the low-level features such as relations among features. Combining the latent information with the original features helps build a robust model in AD/MCI classification, with high diagnostic accuracy. Furthermore, thanks to the unsupervised characteristic of the pre-training in deep learning, we can benefit from the target-unrelated samples to initialize parameters of SAE, thus finding optimal parameters in fine-tuning with the target-related samples, and further enhancing the classification performances across four binary classification problems: AD vs. healthy normal control (HC), MCI vs. HC, AD vs. MCI, and MCI converter (MCI-C) vs. MCI non-converter (MCI-NC). In our experiments on ADNI dataset, we validated the effectiveness of the proposed method, showing the accuracies of 98.8, 90.7, 83.7, and 83.3 % for AD/HC, MCI/HC, AD/MCI, and MCI-C/MCI-NC classification, respectively. We believe that deep learning can shed new light on the neuroimaging data analysis, and our work presented the applicability of this method to brain disease diagnosis.
| S-EPMC4065852 | biostudies-literature
Project description:Deep learning (DL) models typically require large-scale, balanced training data to be robust, generalizable, and effective in the context of healthcare. This has been a major issue for developing DL models for the coronavirus disease 2019 (COVID-19) pandemic, where data are highly class imbalanced. Conventional approaches in DL use cross-entropy loss (CEL), which often suffers from poor margin classification. We show that contrastive loss (CL) improves the performance of CEL, especially in imbalanced electronic health records (EHR) data for COVID-19 analyses. We use a diverse EHR dataset to predict three outcomes: mortality, intubation, and intensive care unit (ICU) transfer in hospitalized COVID-19 patients over multiple time windows. To compare the performance of CEL and CL, models are tested on the full dataset and a restricted dataset. CL models consistently outperform CEL models, with differences ranging from 0.04 to 0.15 for area under the precision and recall curve (AUPRC) and 0.05 to 0.1 for area under the receiver-operating characteristic curve (AUROC).
| S-EPMC8542449 | biostudies-literature
Project description:Here we report single-cell RNA-sequencing generated from one COVID-19 lung removed at the time of organ transplant.
2020-12-29 | GSE163919 | GEO