Big data-based grey forecast mathematical model to evaluate the effect of Escherichia coli infection on patients with lupus nephritis

Abstract

The grey predictive mathematical model based on big data was used for analysis on the effect of Escherichia coli infection on patients with lupus nephritis (LN) in this study. Then, 156 patients diagnosed with LN infections by Wuzhong People's Hospital's information system (HIS) from October 30, 2017 to October 30, 2019 were selected as the experimental group, and 89 patients without LN infections were selected as the control group. Besides, the grey theory mathematical model was applied to process the integrated data, and feature analysis was employed to screen out disease-related bio-markers for the diagnosis of LN. The two groups were compared for affected organs, treatment, laboratory indicators, pathogenic bacteria, and recovery status. Multivariate logistic regression was used to analyze the related factors of patients with infections. The results showed that the specificity, sensitivity, and accuracy of the big data diagnosis based on the grey theory mathematical model were 78.9%, 87.6%, and 92.1, respectively; hormones, c-reactive protein, procalcitonin, and the daily antibiotic dose were positively correlated with concurrent infections (P < 0.05); 38 cases of Gram-negative bacteria were screened out, accounting for the largest proportion (37.18%); the effective rate of the experimental group was obviously lower than that of the control group (P < 0.05), suggesting that C-reactive protein (CRP), procalcitonin (PCT), antibiotics, daily dose of hormones, and serum albumin were independent risk factors for LN infection. In conclusion, the grey predictive mathematical model based on big data had high specificity, sensitivity, and accuracy in diagnosing the occurrence of infection in patients with LN; LN infection was mainly respiratory infection, and gram-negative bacteria were the main pathogen. Patients with LN infections showed higher serum creatinine, 24-hour urine protein quantification, CRP, and PCT, and lower serum albumin and recovery effect versus those without LN infections. © 2021 The Author(s)

Description

This article is licensed under Creative Commons License and full text is openly accessible in CUD Digital Repository. The version of the scholarly record of this article is published in Results in Physics (2021), accessible online through this link https://doi.org/10.1016/j.rinp.2021.104339

Keywords

Bio-marker, Escherichia coli, Gram-negative bacteria, Grey theory prediction, Lupus nephritis

Citation

Fan, M., Gu, S., Jin, Y., Ding, L., Ghonaem, E., & Arbab, A. M. H. (2021). Big data-based Grey Forecast Mathematical Model to Evaluate the Effect of Escherichia Coli Infection on Patients with Lupus Nephritis. Results in Physics, 26, 104339. https://doi.org/10.1016/j.rinp.2021.104339

DOI