37 Using a new technique of ASL | Trigonocephaly Statistical Analyses Statistical analyses were performed using R (Version 4.0.3; R Core Team, 2020).22 The descriptive characteristics are presented as mean and SD or as median and interquartile range (IQR), depending on whether the data are normally distributed or not. Categorical data are presented as counts. Age differences between patients and controls were assessed using Student’s t-test. A chi square test was performed to assess the effect of sex between groups. A significance level of 0.05 was chosen for all tests. Differences in TC between patients and controls were investigated for each of the normalizations (T1w, rigid-body, affine, and DCT) using a scatterplot and a boxplot. In addition, to investigate the differences in the consensus rating of different types of normalization, we fitted a Bayesian proportional odds cumulative logit mixed model. Besides the type of registration, the model included the patient’s group (patient vs. control) and random intercepts for each subject to take into account the fact that different types of normalization performed in the same child are likely correlated. The results are presented as odds ratios and corresponding 95% credible intervals (CI). To facilitate interpretation of the estimated differences in ratings between the different types of normalization, we plotted the estimated probability of receiving a particular rating for each type of normalization (in patients). To compare the mean CBF in three gyri of the frontal lobe in trigonocephaly patients vs. controls for different normalizations, we fitted a mixed linear model for each of the gyri. A random intercept was included to take into account the fact that measurements of different gyri in the same child are likely correlated. Mean regional CBF that was assessed in an anatomical region of interest cannot be reliably assessed in group analyses in children for whom the spatial normalization failed. To avoid bias in the CBF analysis, we have excluded all children from whom one of the four methods failed to generate CBF map in standard space or from whom the map was rated as unusable in the qualitative analysis. Due to the small sample size, no other variables could be included in the model. 2
RkJQdWJsaXNoZXIy MTk4NDMw