ABSTRACT
Heterogeneity and irregularity of multi-source data sets present a significant challenge to time-series analysis. In the literature, the fusion of multi-source time-series has been achieved either by using ensemble learning models which ignore temporal patterns and correlation within features or by defining a fixed-size window to select specific parts of the data sets. On the other hand, many studies have shown major improvement to handle the irregularity of time-series, yet none of these studies has been applied to multi-source data. In this work, we design a novel architecture, PIETS, to model heterogeneous time-series. PIETS has the following characteristics: (1) irregularity encoders for multi-source samples that can leverage all available information and accelerate the convergence of the model;(2) parallelised neural networks to enable flexibility and avoid information over-whelming;and (3) attention mechanism that highlights different information and gives high importance to the most related data. Through extensive experiments on real-world data sets related to COVID-19, we show that the proposed architecture is able to effectively model heterogeneous temporal data and outperforms other state-of-the-art approaches in the prediction task.
ABSTRACT
Objective: To raise awareness about 2019 novel coronavirus pneumonia (NCP) and reduce missed diagnosis rate and misdiagnosis rate by comparing the clinical characteristics between RNA positive and negative patients clinically diagnosed with NCP. Methods: From January 2020 to February 2020, 54 patients who were newly diagnosed with NCP in Wuhan Fourth Hospital were included in this study. RT-PCR method was used to measure the level of 2019-nCov RNA in pharyngeal swab samples of these patients. The patients were divided into RNA positive and negative group, and the differences of clinical, laboratory, and radiological characteristics were compared. Results: There were 31 RNA of 2019-nCov positive cases, and 23 negative cases. Common clinical symptoms of two groups were fever (80.64% vs. 86.96%) , chills (61.29% vs. 52.17%) , cough (80.64% vs. 95.65%) , fatigue (61.30% vs. 56.52%) , chest distress (77.42% vs.73.91%) . Some other symptoms were headache, myalgia, dyspnea, diarrhea, nausea and vomiting. The laboratory and radiological characteristics of two groups mainly were lymphopenia, increased erythrocyte sedimentation rate, increased C-reactive protein, increased lactate dehydrogenase, decreased oxygenation index, normal white blood cell count and bilateral chest CT involvement. There was no statistically significant difference in other clinical characteristics except for dyspnea between two groups. Conclusions: RNA positive and negative NCP patients shared similar clinical symptoms, while RNA positive NCP patients tended to have dyspnea. Therefore, we should improve the understanding of NCP to prevent missed diagnosis and misdiagnosis; In addition, more rapid and accurate NCP diagnostic approaches should be further developed.