Computational Analysis Identifies Novel Biomarkers for High-Risk Bladder Cancer Patients

Timing patterns of initial respiratory syncytial virus infection and factors influencing disease severity in hospitalized infants with different health status

This study aimed to determine the timing patterns of the initial respiratory syncytial virus (RSV) infection and to identify the factors influencing disease severity in infants of varying health status. A retrospective study was conducted at the Affiliated Children's Hospital of Chongqing Medical University from 2012 to 2022. The timing of the first RSV infection was estimated in infants with differing health status using correlation analysis, considering their birth time. Logistic regression was utilized to identify factors influencing severe RSV infection in these infants. RSV detection primarily occurred in the winter and spring. Epidemic season and peak timing of RSV were not significantly affected by health status or the COVID-19 pandemic. A strong positive correlation was observed between the age at RSV infection and the interval from birth to the RSV peak season. Infants born during the RSV epidemic season exhibited a higher likelihood of infection within the first 2 months postbirth. In contrast, those born outside the RSV epidemic season were more susceptible to infection during the subsequent peak. Notably, infants with pre-existing health conditions contracted RSV at an earlier age compared to their healthy counterparts. Among healthy infants, severe RSV infection was associated with sex, age, and timing of infection. For infants with underlying conditions, severe RSV infection was primarily related to age and timing of infection. The initial timing of RSV infection in infants varied depending on their health status. Young age and infection timing during the RSV epidemic season were significant risk factors for severe RSV infection. These findings provide a theoretical basis for optimizing immunization strategies for infants with diverse health conditions.

Screening of neoplastic diseases by statistical analysis of urine fluorescence spectroscopic data. Application of multivariate techniques for enhancing classification

The composition of human fluids is modified during the course of neoplastic diseases. Urine analysis offers the advantage of being a noninvasive method for which samples are easily and routinely collected from patients. In this work, urine fluorescence spectra recorded upon excitation at 405 nm were obtained from healthy volunteers and individuals with different oncologic pathologies. A large number of indexes, i.e., parameters obtained from spectral data which assist spectral features characterization, were developed to classify healthy and pathological populations. The discrimination ability of simple predictive indexes, obtained from spectra pretreated with different normalization procedures and by taking their derivatives, was statistically assessed. In addition, multivariate methods, such as principal component analysis and multivariate curve resolution by alternating least squares, were used to develop more elaborate indexes for distinguishing between healthy and pathological populations. All indexes were systematically evaluated on a statistical basis by in lab-developed routines capable of detecting outliers, judging the normal distribution of the indexes, evaluating variance homogeneity, testing the difference between the means of healthy and pathological populations, as well as performing a receiver operator curve analysis to assess the classification power of each index. Those indexes with the best performances were further combined to perform a linear discriminant analysis, which yielded a powerful classification algorithm with an area under the receiver operator curve of 0.986, a sensitivity of 97.7%, a specificity of 100%, and an overall accuracy of 98.8%. The present study shows that the statistical analysis of urine fluorescence data with a proper combination of multivariate techniques bears a high potential to develop massive screening tests for the early detection of oncologic pathologies.