Novel Non-Invasive Investigation of Spinal Cord Electrical Activity in Health and Neurodegeneration

Abstract

The spinal cord (SC) extends from the lower end of the medulla oblongata to the lumbar region and is divided into five segments: cervical, thoracic, lumbar, sacral, and coccygeal. The spinal cord, as part of the central nervous system, consists of grey and white matter. The spinal grey matter is surrounded by white matter, unlike the brain, where the cortical grey matter makes up the outer layers. The spinal cord is a crucial component of the sensory-motor system and has direct connections with the brain and the periphery, indicating its integral role in sensory-motor communication. The SC has been reported to be affected in neurodegenerative conditions like multiple sclerosis, with the extent of spinal neural degeneration exhibiting a strong correlation with both present and future levels of disability. Despite playing a crucial role in sensory-motor communication, the SC has been largely ignored by studies investigating the neurophysiology/pathophysiology of motor control in health and neurodegeneration. Here, we aim to develop a novel technique for non-invasive recording of high-density (HD) neuroelectric signals from the spinal cord, electrospinalography (ESG), to investigate spinal activity in health and disease. This study features three broad objectives; 1) to design a standardized electrode placement system to record HD-ESG, 2) to investigate spinal neural oscillatory characteristics and spinal functional connectivity with the cortex during motor control in healthy individuals, 3) to investigate the evoked high-frequency oscillations in health and neurodegeneration (Multiple Sclerosis, MS). First, a standardized electrode placement system (SC10X/U) was proposed for recording HD-ESG signals to investigate the spinal neuroelectric activity at the cervical and upper thoracic vertebral levels. This addresses the technical gaps; despite several recent studies utilizing high-density electrode montages to record spinal activity in response to peripheral stimulation, no electrode placement system exists to facilitate standardized recording across individuals and research groups. The proposed SC10X/U system can incorporate up to 76 electrodes, with their locations standardized w.r.t sixteen distinct surface anatomical markers. The system divided the total medial-lateral distance in equal proportions of 10%, and the total medial-sagittal distance was divided in proportion to the inter-vertebral spinous distances. The electrode space was symmetrically divided on either side of the spinal cord, and a labelling nomenclature featuring two uppercase letters followed by a number was proposed for identifying the unique 76 locations. Second, HD-ESG was recorded in conjunction with HD-EEG to investigate spinal neural activity characteristics and cortico-spinal functional connectivity during voluntary isometric motor task. To our knowledge, this is the first study to non-invasively evaluate the spatio-spectral characteristics of spinal neural activity and its coherence with brain activity. Frequency-dependent neural oscillations and their neurophysiology have been widely reported in the cortex. Here, we demonstrated that spinal neurophysiology also exhibits frequency-dependent neural activity. Spinal signals in the beta-band demonstrated significant ipsi-anterolateralized (p<0.05) activity that remained activate for the sustained part of the contraction. Furthermore, ipsi-anterolateral spinal electrodes indicated significant coherence (p<0.05) with the contra-lateral motor cortex, indicating its involvement in descending cortico-spinal control. Unlike the beta-band, no significant ipsi-lateralization was observed for low-gamma spinal activity during the sustained period of contraction and bi-lateral activity was observed in the high-gamma band. In the low-gamma band, dorsal midline electrodes demonstrated increased spinal activity. A significant coherence (p<0.05) between the spinal dorsal-midline region and central-midline EEG electrode was also observed in the low-gamma band. Additionally, ipsi-anterolateral electrodes demonstrated a steady increase in low-gamma band activity over time during the sustained phase. These findings indicate the inherent role of frequency-dependent neural oscillations in both the SC and the cortex during isometric motor control. Third, ipsilaterally localised evoked high-frequency oscillations (HFOs) in the 400-600 Hz wide-frequency band were observed in response to the median nerve stimulation for both the control and MS participant groups. The onset latency did not show any significant difference between the two groups. The evoked HFOs at the ipsilateral electrode position, C6 vertebral level, lasted for 7.2 ± 1.5 ms and 6.7 ± 0.75 ms for the MS and control group respectively. The duration of the HFOs activity did not show significant differences between the two groups. A significant difference in the peak amplitude of the HFOs envelope was observed between the two groups for the ipsilateral electrode at the C6 vertebral level (p=0.021), with people with MS demonstrating significantly lower amplitude than controls. Interestingly, the MS group had decreased spectral power in the 400-600 Hz band for the duration of the evoked-HFOs, indicating a potential disruption in the synchronization of the evoked high-frequency oscillation. In conclusion, the proposed HD-ESG system has the potential to uncover the role of spinal neurophysiology in sensory-motor communication by non-invasive recording of the spinal neuro-electric activity during external stimulation and/or during rest/tasks (resting-state, voluntary motor task). Here, we demonstrated that spinal neuroelectric activity/responses can be utilized to evaluate the frequency-band dependent oscillatory characteristics and cortico-spinal functional connectivity. In future, the aim is to extend the current analysis and investigate the connectivity between brain, spinal cord, and periphery during motor control. Furthermore, we aim to assess the utility of spinal neuro-oscillatory characteristics in neurodegeneration as a potential marker.