Searcharxiv⌕ Search

arXiv subjects

Jeffrey P. Guenette

Publications and source records attributed to Jeffrey P. Guenette.

2 recordsLinked to original sources

Unveiling neuronal microstructure in the human brain in vivo with time-dependent radial diffusivity in MRI

Diffusion time-dependence, defined as variations in diffusivity and/or diffusional kurtosis with diffusion time, has emerged as a valuable non-invasive imaging marker for characterizing tissue microstructural features, such as cell size, density, packing disorder, and membrane permeability. In white matter, diffusion time-dependent changes between the short diffusion time and long diffusion time in radial diffusivity (RD), defined as the diffusivity perpendicular to fiber tracts, were demonstrated to correlate strongly with mean axon diameter in ex vivo spinal cord tissues, and to reveal demyelination in mouse corpus callosum. Despite their potential to non-invasively unveil neuronal microstructures to improve the assessment and targeted therapy of neurological diseases, these novel image contrasts obtained at short diffusion times using oscillating gradient spin echo (OGSE) have only recently become feasible for human in vivo studies with high-performance gradient MRI systems. In this preliminary study, we characterized time-dependent RD with OGSE encoding in the human brain in vivo. The change in radial diffusivity between short diffusion time and long diffusion time (delta_RD) consistently exhibited high values in the corticospinal tract, indicating high sensitivity of delta_RD to large axon diameter in human brains. Imaging at a high OGSE frequency of 100 Hz and a moderate b-value of 800 s/mm2 produced the highest delta_RD in the corticospinal tract. This study established a baseline for future investigations of neuronal microstructural alterations in neurological disorders and diseases.

physics.med-ph↗

Registration of Longitudinal Spine CTs for Monitoring Lesion Growth

Accurate and reliable registration of longitudinal spine images is essential for assessment of disease progression and surgical outcome. Implementing a fully automatic and robust registration is crucial for clinical use, however, it is challenging due to substantial change in shape and appearance due to lesions. In this paper we present a novel method to automatically align longitudinal spine CTs and accurately assess lesion progression. Our method follows a two-step pipeline where vertebrae are first automatically localized, labeled and 3D surfaces are generated using a deep learning model, then longitudinally aligned using a Gaussian mixture model surface registration. We tested our approach on 37 vertebrae, from 5 patients, with baseline CTs and 3, 6, and 12 months follow-ups leading to 111 registrations. Our experiment showed accurate registration with an average Hausdorff distance of 0.65 mm and average Dice score of 0.92.

eess.IV↗