DKI enhances the sensitivity and interpretability of age-related DTI patterns in the white matter of UK biobank participants

• Published in: Neurobiology of aging 2022-07, Vol.115, p.39-49
• Main Authors: Taha, Hiba T., Chad, Jordan A., Chen, J. Jean
• Format: Article
• Language: English
• Subjects: Aged; Biological Specimen Banks; Brain - diagnostic imaging; Brain aging; Crossing fibres; Diffusion kurtosis imaging (DKI); Diffusion tensor imaging (DTI); Diffusion Tensor Imaging - methods; Humans; Neurodegeneration; UK biobank; United Kingdom; White matter; White Matter - diagnostic imaging
• ISSN: 0197-4580; 1558-1497
• DOI: 10.1016/j.neurobiolaging.2022.03.008

•We investigate how DKI can enhance DTI findings of aging adult white matter.•Diffusivity values computed from DKI are 15%–20% greater than those computed from DTI.•A greater extent of WM shows kurtosis than diffusivity age associations (75% vs. 66%).•Overlapping diffusivity/kurtosis patterns in DKI help form biological interpretation. Studies of healthy brain aging traditionally report diffusivity patterns associated with white matter degeneration using diffusion tensor imaging (DTI), which assumes that diffusion measured at typical b-values (approximately 1000 s/mm2) is Gaussian. Diffusion kurtosis imaging (DKI) is an extension of DTI that measures non-Gaussian diffusion (kurtosis) to better capture microenvironmental processes by incorporating additional data at a higher b-value. In this study, using diffusion data (b-values of 1000 and 2000 s/mm2) from 700 UK Biobank participants aged 46–80, we investigate (1) the extent of novel information gained from adding diffusional kurtosis to diffusivity observations in aging, and (2) how conventional DTI metrics in aging compare with diffusivity metrics derived from DKI, which are corrected for kurtosis. We establish a pattern of lower kurtosis alongside higher diffusivity among older adults, with kurtosis generally being more sensitive to age than diffusivity. We also find discrepancies between diffusivity metrics derived from DTI and DKI, emphasizing the importance of accounting for non-Gaussian diffusion when interpreting age-related diffusivity patterns.