R language version 4

Categorical data are shown to be the frequency with proportion and explored based on Chi-square test or Fisher’s exact test. With the aim of calculating the PFS and OS and plot the curve, the Kaplan-Meier method was employed. The log-rank test was adopted for comparing the two groups. A 2-tailed p-value ≤0.05 represented statistical significance. Cox proportional hazards models were applied, aiming to explore the correlation between the covariates and PFS or OS. Variables showing p < 0.05 in univariate analysis were subjected to stepwise multivariate analysis. Moreover, all data calculations were conducted by employing R language version 4.0.4 (R Foundation for Statistical Computing, Vienna, Austria).

Statistical analyses were performed using R language, version 4.1.0 (R Foundation for Statistical Computing, Vienna, Austria). The R packages doBy and ggplot2 were applied. All continuous variables are expressed as means ± standard deviation (SD), while categorical variables are shown as totals and proportions. The interobserver consistencies for all quantitative parameters between the two radiologists were evaluated with the intraclass correlation coefficient (ICC) analysis. Agreement was defined as good (ICC > 0.75), moderate (ICC = 0.5–0.75), or poor (ICC < 0.5). Clinicopathological and DCE-MRI features were compared using the Mann–Whitney U and chi-square tests. Quantitative parameters (T1, T2, PD, and ADC values) were performed both univariate and multivariate logistic regression analyses with a variable selection criterion of p < 0.05. We estimated the area under the receiver operating characteristic curve (AUC) to evaluate the predictive ability of the quantitative parameters. For all tests, p < 0.05 was considered statistically significant.

We evaluated cross-sectional associations between environmental exposures (greenspace, ALAN, noise, and air pollution) and self-reported sleep duration, sleep latency (time to fall asleep), and trouble staying asleep using logistic regression, calculating odds ratios (ORs) and 95% confidence intervals (95% CI). We mutually adjusted for all environmental exposures and a priori covariates, race/ethnicity (Asian, Black, Hispanic, Mixed, White, or Unknown/Missing), sex (male vs. female), age, SES (maternal education less than college vs. college graduate), and CHS community.^{7 (link),10 (link),13 (link),18 (link),38 (link),39 (link)} Associations were reported per interquartile range (IQR) change in exposure. Because these environmental exposures were previously associated with stress in our study population,^{26 (link)} we assessed the mediating role of self-reported stress on our exposure-sleep associations and report it as a percent.⁴⁰ Due to possible interactions between environmental exposures and SES,^{41 (link),42 (link)} we assessed interaction between each exposure and SES among those that reported maternal education, further stratifying for those that were statistically significant (P-value < 0.05). All analyses were conducted in the R Language, version 4.1.0 (R Foundation for Statistical Computing, Vienna, Austria).

The x2 test evaluated differences and relationships between categorical parameters. Survival analysis was performed considering as outcomes the overall survival (OS), defined as the time between the date of treatment and the date of death; disease-free survival (DFS), defined as the time between the date of treatment and recurrence; and neck control and metastases-free survival (NC-MFS), as defined as a regional or distant relapse of the disease. We evaluated the impact of the different types of neck dissections on the aforementioned oncological outcomes. We first assessed the whole cohort of patients; then, we further analyzed the subgroups of patients affected by multiple nodal disease and extranodal extension (ENE) because these are the subcategory believed to benefit from more extended neck treatment. Survival curves were plotted by the Kaplan–Meier method, and the log-rank test compared the differences between curves. To reduce the effect of confounder values in OS and DFS, with conducted a multivariable analysis with the Cox regression model. Statistical analyses were performed by SPSS 18.0 software (SPSS Inc., Chicago, IL, USA), and p < 0.05 was considered statistically significant. The survival curves were constructed with packages of “survival” and “survminer” in R language version 4.1.0 (R Foundation for Statistical Computing, Vienna, Austria) [30 ,31 ,32 (link),33 ].

All radiomics procedures and statistical analyses were conducted with Python software (Version 3.8.5). R language (Version 4.2.2, R Foundation for Statistical Computing, Vienna, Austria) was used for waterfall plots, calibration curves, and DCA. Continuous variables were compared using t-test or Mann-Whitney U test, and categorical variables were compared with chi-square test. P values less than 0.05 were regarded as statistically significant.