Analysis of Risk Factors for Urinary Tract Infection in Children and Construction and Validation of a Prediction Model.

OBJECTIVE: To analyse the risk factors for urinary tract infection (UTI) in children and construct and validate a risk prediction model. METHODS: The study selected 258 children with suspected UTI in the paediatric department of our hospital from August 2019 to August 2021. Identified as the subjects in this research, paediatric patients' clinical data were used for retrospective analysis. Based on the counting results of urinary leucocytes and bacteria, children were divided into the UTI group (n = 67) and non-UTI group (n = 191). Univariate analysis and multivariate logistic regression analysis were used to screen the independent risk factors for UTI in children, and a prediction model was constructed according to the results. The Hosmer-Lemeshow goodness-of-fit (GOF) test and receiver operator characteristic (ROC) curve analysis were used to validate the calibration and application value of prediction model. RESULTS: Logistic regression analysis identified length of hospitalisation ≥10 days (OR = 3.611, 95% CI: 1.781-7.325), indwelling ureter (odds ratio (OR) = 3.203, 95% CI: 1.615-6.349), history of infection (OR = 4.827, 95% CI: 2.424-9.612), congenital malformation/dysplasia (OR = 4.212, 95% CI: 2.079-8.531), constipation (OR = 4.021, 95% CI: 1.315-12.299) and anaemia (OR = 2.275, 95% CI: 1.236-4.186) as risk factors for UTI in children (p < 0.05). The formulation method was adopted to construct the following prediction model of UTI in children: Z = 2.066 × (length of hospitalisation ≥10 days) + 1.164 × (indwelling ureter) + 1.574 × (history of infection) + 1.438 × (congenital malformation/dysplasia) + 1.392 × (constipation) + 0.882 × (anaemia). The test results revealed the good GOF and high calibration (χ2 = 9.077, p = 0.336) of prediction model. Furthermore, the area under the ROC curve was 0.825 (95% CI: 0.766-0.884, p < 0.001), indicating the good discrimination and prediction efficiency of model. CONCLUSIONS: Based on clinical results, further attention should be given to high-risk children with UTI, and intervention measures should be taken immediately. The application and popularisation of prediction model will allow us to provide strategic guidance for preventing and treating UTIs in clinics.

The Optimally Designed Variational Autoencoder Networks for Clustering and Recovery of Incomplete Multimedia Data.

Clustering analysis of massive data in wireless multimedia sensor networks (WMSN) has become a hot topic. However, most data clustering algorithms have difficulty in obtaining latent nonlinear correlations of data features, resulting in a low clustering accuracy. In addition, it is difficult to extract features from missing or corrupted data, so incomplete data are widely used in practical work. In this paper, the optimally designed variational autoencoder networks is proposed for extracting features of incomplete data and using high-order fuzzy c-means algorithm (HOFCM) to improve cluster performance of incomplete data. Specifically, the feature extraction model is improved by using variational autoencoder to learn the feature of incomplete data. To capture nonlinear correlations in different heterogeneous data patterns, tensor based fuzzy c-means algorithm is used to cluster low-dimensional features. The tensor distance is used as the distance measure to capture the unknown correlations of data as much as possible. Finally, in the case that the clustering results are obtained, the missing data can be restored by using the low-dimensional features. Experiments on real datasets show that the proposed algorithm not only can improve the clustering performance of incomplete data effectively, but also can fill in missing features and get better data reconstruction results.