Resting state fMRI investigation on spinal cord and cervical myelopathy

Abstract

Resting state fMRI (rsfMRI) is a widely used neuroimaging technique which focuses on the intrinsic functional neural activity of the brain. As another essential part of the central nervous system, the spinal cord, which is anatomically and physiologically connected to the brain, has not been well investigated by rsfMRI. The rsfMRI provides a novel in-vivo investigation technique of central nervous system, which could be applied for the spinal cord diseases. However, technical challenges block the application of spinal cord rsfMRI, which include the small size of the spinal cord, magnetic inhomogeneity, and physiology noise, etc. To develop the spinal cord rsfMRI technique, those aforementioned technical challenges must be overcome. Recent studies have adopted adjustments to rsfMRI image acquisition to successfully solve problems such as small size of spinal cord and susceptibility artifact. One of the key issue remaining is the non-ignorable physiology noise. In this study, an ICA-based nuisance regression (ICANR) method was developed to reduce the physiology noise. Experimental results showed that ICANR could effectively reduce physiology noise and improve the robustness of rsfMRI investigation on the spinal cord. It enables more accurate investigation of spinal cord intrinsic functional connectivity. After the technical development, the spinal cord rsfMRI technique was applied on normal subjects to investigate the intrinsic functional neural activity. Functional connectivity analysis found that the topology of the functional network was not uniformly distributed across cervical vertebral levels. Networks in the upper cervical cord presented a high level of connectivity. The graph theory analysis of resting state network properties proves that the intrinsic functional activity in the cervical spinal cord is efficient both globally and locally. Through the investigation of spinal cord rsfMRI, the organization of the intrinsic functional network in the cervical spinal cord is characterized by an efficient well-organized network in which the upper cervical cord exhibits higher degrees of connectivity and serves as a hub within the functional network. Besides, the spinal cord rsfMRI technique was also used to explore the neural development during spinal cord disease. Cervical myelopathy (CM) patients were investigated. Functional connectivity and amplitude of low frequency fluctuation analysis demonstrated the increase of intrinsic functional neural activity in CM patients, indicating the functional neural plasticity of the spinal cord after CM. These findings may help understand the natural progression of dysfunction after CM. In addition, brain rsfMRI investigation was performed for CM patients in order to investigate the functional alteration of brain during CM. The graph theory analysis revealed the increase of cerebellum local efficiency in CM patients. It may deepen our understanding of the brain neural development under CM. In summary, this study developed a method that enable the rsfMRI application to the spinal cord. The rsfMRI investigation of cervical cord and CM revealed the neural functional plasticity on the spinal cord as well as the functional reaction on the cerebellum.