Effects of Acquisition Protocol on Advanced Diffusion MRI Measures. (20th December 2022)

• Record Type: Journal Article
• Authors: Mishra, Virendra R; Sreenivasan, Karthik; Cordes, Dietmar; Ritter, Aaron
• Is Part Of: Alzheimer's & dementia. Volume 18(2022)Supplement 1
• Page Start: n/a
• Page End: n/a
• Publication Date: 2022-12-20
• Subjects: Alzheimer's disease -- Periodicals; Alzheimer Disease -- Periodicals; Dementia -- Periodicals; Démence; Maladie d'Alzheimer; Périodique électronique (Descripteur de forme); Ressource Internet (Descripteur de forme); 616.83
• DOI: 10.1002/alz.066683
• Languages: English
• ISSNs: 1552-5260

Abstract

Abstract: Background: Several studies have suggested that advanced diffusion MRI (dMRI) measures derived through the analytic techniques of diffusion kurtosis imaging (DKI) and neurite orientation density and dispersion imaging (NODDI) might provide better sensitivity to predict the progression of Alzheimer's disease (AD) when compared to conventional single tensor diffusion tensor imaging. Two popular simultaneous multislice (SMS) sequences, namely Alzheimer's Disease Neuroimaging Initiative (ADNI)‐3 and Human Connectome Protocol (HCP) are currently widely used in the acquisition of such advanced dMRI data. Since HCP sequences require a custom connectome scanner, several variants of HCP protocol compatible with clinical MRI are in wide use. Before such techniques are used widely across the globe to understand AD, it is imperative to understand whether the absolute WM measures derived through such advanced dMRI measures are reproducible and whether harmonization of dMRI across sites employing different acquisition protocols is applicable. In this study, we hypothesize that higher spatial and angular resolution dMRI data will result in higher quantification of DKI and NODDI metrics, as compared with comparatively lower spatial resolution advanced ADNI‐3 sequences. Method: A 32‐year‐old healthy male participant was scanned over five weeks, both with Advanced ADNI‐3 protocol (114 directions, 2mm 3 isotropic, b‐values=500s/mm 2, 1000s/mm 2, and 2000s/mm 2 ) and 2 variants of the HCP sequence (1.5mm 3 and 2mm 3 ) both with 3 b‐values (500s/mm 2, 1000s/mm 2, and 2500s/mm 2 ) and 213 directions on a 3T Skyra scanner. Total scan time for the ADNI‐3 sequence was 11:37 minutes, 2mm 3 CCF sequence was 18:41 minutes, and 1.5mm 3 CCF sequence was 23:13 minutes. Various DKI and NODDI measures were computed on all repetitions in‐house. JHU atlas of WM tracts was used as a mask to extract values along each tract at each run separately. Result: As shown in Figs 1‐4, dMRI‐derived measures were underestimated for most of the WM tracts for the ADNI‐3 sequence, although they had less variance across different runs. Second, NODDI measures were more reproducible across most WM tracts when compared to DKI measures across protocols. Conclusion: Acquisition parameters affect the absolute quantification of advanced dMRI measures. Our data suggest that the pooling of dMRI data across various protocols should be performed with caution