Discrimination of errors from neuronal activity in functional MRI of the human spinal cord by means of general linear model analysis

Abstract

Functional MRI (fMRI) of the spinal cord has been demonstrated to provide reliable and sensitive maps of neuronal activity, particularly when combined across several experiments. Individual experiments reveal neuronal activity as well as errors. The dominant source of errors is hypothesized to be physiological motion, including cardiac and respiratory motion, flow of blood and cerebrospinal fluid (CSF), and motion of the spinal cord within the spinal canal. All of the hypothesized sources of error are therefore related to cardiac and respiratory motion, which can be recorded during an fMRI experiment. Analyses were carried out with a general linear model (GLM) with peripheral pulse and respiration recordings used as models of errors. The results demonstrate that the sensitivity of spinal fMRI is improved and errors are reduced when peripheral pulse traces are used in the GLM, but no improvement was detected with the inclusion of respiratory traces.