Diffusional kurtosis imaging (DKI) incorporation into an intravoxel incoherent motion (IVIM) MR model to measure cerebral hypoperfusion induced by hyperventilation challenge in healthy subjects

• Published in: Magma (New York, N.Y.) Vol. 30; no. 6; pp. 545 - 554
• Main Authors: Pavilla, Aude, Gambarota, Giulio, Arrigo, Alessandro, Mejdoubi, Mehdi, Duvauferrier, Régis, Saint-Jalmes, Hervé
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.12.2017; Springer Verlag
• Subjects: Bioengineering; Biomedical Engineering and Bioengineering; Computer Appl. in Life Sciences; Engineering Sciences; Health Informatics; Imaging; Life Sciences; Medicine; Medicine & Public Health; Radiology; Research Article; Signal and Image processing; Solid State Physics; Kurtosis; Diffusion magnetic resonance imaging; Perfusion; Hyperventilation; Intravoxel incoherent motion
• ISSN: 0968-5243; 1352-8661; 1352-8661
• DOI: 10.1007/s10334-017-0629-9

Objectives The objectives were to investigate the diffusional kurtosis imaging (DKI) incorporation into the intravoxel incoherent motion (IVIM) model for measurements of cerebral hypoperfusion in healthy subjects. Materials and methods Eight healthy subjects underwent a hyperventilation challenge with a 4-min diffusion weighted imaging protocol, using 8 b values chosen with the Cramer-Rao Lower Bound optimization approach. Four regions of interest in gray matter (GM) were analyzed with the DKI–IVIM model and the bi-exponential IVIM model, for normoventilation and hyperventilation conditions. Results A significant reduction in the perfusion fraction ( f ) and in the product fD * of the perfusion fraction with the pseudodiffusion coefficient ( D *) was found with the DKI–IVIM model, during the hyperventilation challenge. In the cerebellum GM, the percentage changes were f : −43.7 ± 40.1, p  = 0.011 and fD *: −50.6 ± 32.1, p  = 0.011; in thalamus GM, f : −47.7 ± 34.7, p  = 0.012 and fD *: −47.2 ± 48.7, p  = 0.040. In comparison, using the bi-exponential IVIM model, only a significant decrease in the parameter fD * was observed for the same regions of interest. In frontal-GM and posterior-GM, the reduction in f and fD * did not reach statistical significance, either with DKI–IVIM or the bi-exponential IVIM model. Conclusion When compared to the bi-exponential IVIM model, the DKI–IVIM model displays a higher sensitivity to detect changes in perfusion induced by the hyperventilation condition.