Introduction: The ongoing pandemic due to coronavirus disease 2019 has put tremendous strain on the healthcare system around the world. There is a paucity of data describing the role of National Early Warning Score 2 (NEWS2) in the assessment of COVID-19 cases. This study aimed at identifying NEWS2 calculated on admission as a valuable tool for risk stratification and prediction of in-hospital mortality in COVID-19 patients.Materials and method: This prospective, observational study included 814 confirmed COVID-19 cases and was conducted over a period of three months. Vital parameters were assessed and NEWS2 was calculated on admission. Data were entered in excel format and statistical analysis was done in Python 3.8 statistical software (Wilmington, DE: Python Software Foundation). Pearson's chi-squared test was used following which a significant NEWS2 cut-off score to predict in-hospital mortality was determined by means of receiver operating characteristic (ROC) curve.Results: Mortality of 9.09% was noted and correlations were made with age, comorbidity, and NEWS2 score. For in-hospital deaths, comorbidities were present in 66.21% of patients, the mean age was 60.14 years, and average NEWS2 score was 9. For discharged patients only 21.89% had comorbidities, mean age was 42.96 years, and average NEWS2 score was 1.17. NEWS2 score of ≥ 6 had a sensitivity of 93.24% and specificity of 98.91%, and hence was a statistically significant cut-off value for predicting mortality on admission. Conclusion: Age, presence of comorbidities, and NEWS2 have a positive correlation with mortality in COVID-19 patients. NEWS2 score being easy, reliable, and quick to calculate, should be used to triage these patients on admission. Scores ≥ 6 should be considered to have a higher risk of adverse outcomes and hence should be managed prudently along with clinical judgment.

We present an optical compute engine with implementation of Deep CNNs. CNNs are designed in an organized and hierarchical manner and their convolutional layers, subsampling layers alternate with each other, thus the intricacy of the data per layer escalates as we traverse in the layered structure, which gives us more efficient results when dealing with complex data sets and computations. CNNs are realised in a distinctive way and vary from other neural networks in how their convolutional and subsampling layers are organised. DCNNs bring us very proficient results when it comes to image classification tasks. Recently, we have understood that generalization is more important when compared to the neural network's depth for more optimised image classification. Our feature extractors are learned in an unsupervised way, hence the results get more precise after every backpropagation and error correction.

No abstract