Imputation techniques for the detection of microstructural changes in schizophrenia, with an application to magnetization transfer imaging.

Neuroimaging techniques such as magnetization transfer imaging allow the detection of microstructural alterations of tissue, and for this reason have been applied to the study of disorders such as schizophrenia. However, they are also sensitive to partial volume effects arising from mixed compartments, such as those comprising cerebral spinal fluid, which makes separate evaluation of volumetric and structural alterations difficult. Ensuing regional differences in the distribution of data and signal-to-noise ratio add further potential bias to their assessment. In the present study we simultaneously applied tissue segmentation, statistical imputation, and nonparametric inference to address these issues and improve the validity of statistical inference. In a case study of N=32 schizophrenic patients matched to the same number of controls, we compared a standard voxel-based analysis with one supplemented by the imputation technique. We were able to replicate significant results in the imputed analysis and even extend them in the areas not excluded by excessive partial volume effects. Application of segmentation algorithms in this dataset also suggested that partial volume effects from spinal fluid potentially affect inference in most cortical gray matter, unless remedial steps are undertaken. Refined imputation methods may be particularly attractive in future research settings characterized by large samples and the availability of adequate computational resources.

Components of variance in brain perfusion and the design of studies of individual differences: the baseline study.

Simple baseline studies correlate average perfusion levels measured at rest with individual variables, or contrast subject groups as in case-control studies. In this methodological work, we summarize some formal properties of the design of these studies, and investigate the sources of variance that characterize data acquired with the arterial spin labeling technique, with the purpose of alerting users to the main sources of variation that determine background variance and affect the power of statistical tests. This design typology is characterized by two variance components: between acquisitions and between subjects. We show that variation between acquisitions is affected by the presence of large vessels and venous sinuses, with potential adverse effects especially in the temporal and insular regions, and provide maps of the number of acquisitions or subjects required to reach the desired estimate precision. Furthermore, we show that the largest source of variation between subjects is captured by global perfusion levels, and can in principle be removed by adjusting the data. Significance levels, however, are not always only improved by the adjustment procedure; we provide an example in the correlation with age, and attempt to explain the consequences of the adjustment with the help of a principal component analysis of the data. We also show the existence of variation between subjects in the perfusion in the territory of the posterior cerebral artery and in hemispheric asymmetry.