Abstract
PURPOSE: Few studies have investigated insomnia using intravoxel incoherent motion imaging. The aim ofthis study was to evaluate the classification performance of intravoxel incoherent motion diffusion-weighted imaging in differentiating between insomnia disorder patients and normal control subjects. The potential effects of microstructure and microcirculation between sleep quality and glymphatic system function were subsequently explored. METHOD: Forty-two patients with insomnia disorder and 42 normal controls were included in this study. The demographic characteristics, Pittsburgh Sleep Quality Index scores, and imaging data of the subjects were collected. Magnetic resonance imaging revealed enlarged perivascular spaces that indirectly reflect the glymphatic system. Imaging data were subjected to postprocessing, and parameters including the pure diffusion coefficient, pseudo diffusion coefficient and perfusion fraction were calculated. Partial correlation analysis and mediation analysis were used to explore the role of microcirculation between sleep quality and the perivascular space. FINDINGS: There were brain region-specific differences between two groups, mainly involving perfusion parameters. D* and f in the dorsal-lateral superior frontal gyrus, hippocampus, caudate nucleus, lenticular nucleus, thalamus, amygdala, and fusiform gyrus demonstrated significant differences compared to the healthy group. The support vector machine classifier based on intravoxel incoherent motion imaging parameters has good diagnostic efficiency, with an area under the curve of 0.92. In the mediation analysis, D* in the left thalamus mediated the relationship between sleep quality and enlarged perivascular spaces severity in the centrum semiovale. CONCLUSION: These findings suggest that intravoxel incoherent motion imaging may provide important value for an accurate classification of insomnia. Cerebral microcirculation may have a mediating effect on insomnia and glymphatic system dysfunction.