Evaluation of the association between MREPT-derived conductivity and IVIM-derived ISF-related metrics in the brains of patients with cognitive impairments.
Authors
Affiliations (4)
Affiliations (4)
- Department of Electronics and Information Convergence Engineering, College of Electronics and Information, Kyung Hee University, Yongin-Si, Gyeonggi-do, Republic of Korea.
- MR Clinical Science, Philips Healthcare, Seoul, Republic of Korea.
- Department of Neurology, Kyung Hee University Hospital at Gangdong, Kyung Hee University College of Medicine, Seoul, Republic of Korea.
- Department of Radiology, Kyung Hee University Hospital at Gangdong, Kyung Hee University College of Medicine, Seoul, Republic of Korea.
Abstract
Alterations in the ionic microenvironment and interstitial fluid (ISF) dynamics are increasingly recognized as early pathological events in Alzheimer's disease (AD). However, the <i>in vivo</i> relationship between bulk tissue electrical properties and ISF mobility remains unclear. This study aimed to investigate the relationship between magnetic resonance electrical properties tomography (MREPT)-derived high-frequency conductivity (HFC) and intravoxel incoherent motion (IVIM)-derived ISF-related metrics in patients with cognitive impairments across the cognitively normal to AD spectrum. A total of 161 participants (26 cognitively normal (CN), 78 with amnestic mild cognitive impairment (MCI), and 57 with AD) underwent 3.0 T MRI. HFC maps were reconstructed using phase-based MREPT. ISF-related diffusion (D<sub>ISF</sub>) and volume fraction (F<sub>ISF</sub>) indices were extracted from multi-b-value IVIM data using an unsupervised physics-informed neural network (PINN) approach. Voxel-based and region-of-interest (ROI) statistical analyses, including ANCOVA, partial correlations, and robust multiple regressions, were performed with adjustments for age and sex. Along the disease progression from CN to AD, D<sub>ISF</sub> significantly decreased, while F<sub>ISF</sub> and HFC increased. Partial correlation analyses revealed that these alterations were significantly correlated with age and cognitive decline. Furthermore, MREPT-derived HFC exhibited region-specific correlations with IVIM-derived ISF-related measures, most notably in the bilateral corpus callosum and thalamus. Robust multiple regression analyses revealed limited, region-specific independent associations between ISF indices and HFC, with the most robust finding observed for D<sub>ISF</sub> in the left thalamus, whereas F<sub>ISF</sub>-related associations did not survive correction, and F<sub>PAR</sub> showed the most consistent independent association in the volume-fraction model. The region-specific associations observed between MREPT-derived HFC and IVIM-derived ISF-related metrics in select AD-vulnerable regions, particularly the independent D<sub>ISF</sub>-HFC association in the left thalamus, suggest that neurodegeneration may be linked with local ionic and microenvironmental compartmental shifts. These findings require replication in larger, independent cohorts. Combining MREPT and IVIM-based ISF-related metrics may provide a non-invasive imaging framework to probe these complex microenvironmental alterations in cognitive impairment.