Contribution to the development of spinal functional magnetic resonance imaging as a tool in the investigation of spinal cord physiology

It is important, in order to further establish spinal fMRI as a valuable clinical and research tool, to expand the repeÍoire of stimuli and responses that can be assessed by this method.The specific contributions to the development ofspinal functional magnetic resonance imaging were carried out in four studies.The first study aimed to develop a Iower limb movement task suitable for functional imaging.A pedal was designed, built and tested, and healthy human volunteers paÍicipated in alterrating flexion and extension ankle movements during a single-shot fast spin-echo imaging sequence.Active and passive pedaling was performed by all volunteers.Images werc found to be sufficiently unaffected by motion.Neuronal activity was detected in the do¡sal and ventral homs bilaterally in both conditions, however there was less activity overall in response to passive pedaling.Active and passive pedaling each incuned signal changes of approximately 12Vo.The second study aimed to improve the volume coverage of the spinal cord with increased resolution.This was achieved with a sagittal orientation imaging method of the cervical spinal cord during themal stimulation.The third study employed the new imaging technique for the lumbar spinal cord and the previously tested movement task was carried out with a spinal cord injured population.Neuronal activity was detected in the lumbar cord caudal to the injury site in all injured volunteers.Active pedaling incurred more neu.onalactivity than passive pedaling, similar to the healthy volunteers studied in part one ofthe thesis.signal intensity changes of 13.69o and, 15.\vo were recorded for active and passive participation, respectively.The fourth study of the thesis aimed to discdminate true neuronal activity from false positive activity in a spinal functional imaging study by means of a cluster analysis.It was shown that true neuì.onalactivity-related signal changes that occur in the gray matter tend to occur in the 57o-l5vo signal intensity range, and that apparent activation sumounding the cord tended to occur in the above l5%o signal intensity range.The overall goal of contributing to the development and advancement of spinal fMRI towards reliable research and practical clinical use was achieved. General IntroductionAlthough MRI has been used to anatomically image the spinal cord fo¡ a number of years, it has only recently been successful in investigating functional processes.The purpose of this disse¡tation is to contribute to the development of spinal fMRI as a tool for investigating spinal cord physiology.In order to reach this goal, a number of studies were conducted.First, a spinal fMRI study was carried out that involved healthy volunteels imaged while participating in lower limb movement tasks.Second, we have adjusted the method to obtain images in the sagittal orientation.The numerous difficulties with developing this technique ale outlined in the introduction.Third, once imaging in the sagittal orientation was established in the cervical cord, we used the technique to image the lumbar cord of spinal cord injuled volunteers employing the same tasks used with the healthy volunteers, both to further validate the sagittal imaging method and also to see if spinal fMRI could detect neuronal function caudal to an injury site.Fourth, a paper considering the implications of a cluster analysis of this data ends the manuscript component of this work.In or der to appreciate the unique contribution this project brings to the literature, a review of the MR basics and fMRI studies that lead up to this body of work are outlined, and the relevant spinal cord physiology is provided to assist in the comprehension ofthe imaging results.The four manuscripts are then included in the dissertation to describe the rationale, methods, results and discussion of the above mentioned studies.Finally, an overall discussion concludes the dissertation.present, the orientation of a magnetic moment is random due to its thermal motion.A magnetic moment in a static magnetic field can be in only one of two possible states, either aligned parallel to the fietd or against it (anti-parallel).Each state has a specific energy.When placed in the magnst, after a few seconds the magnetic moments achieve equilibrium where they either align parallel with the static magnetic field or anti-parallel to it.Since the energy of the parallel state is the lower of the two, more magnetic moments are in this state, according to MR convention.It is the net magnetization of the sample, the difference between the number of parallel and anti-parallel magnetic moments, that is observed.The "magnetization" is the net magnetic moment per unit volume.The magnetic field is the static field of the MR system and is called Bs (with the z axis being defined as parallel to Bo).The magnetization depends on the magnetic field strength and the temperature.Increasing B¡ increases the number of protons that align parallel to Bo, as opposed to anti-parallel.Increasing temperature incr.eases the random forces that tend to push the magnetic moments out of alignment into a more t.andomorjentation.The equilibrium magnetization, M6, is the net total magnetic field of the magnetic moments.This magnetization is aligned parallel to the Bo field, is parallel to the z axis, and is zero in the transverse (xy) plane.When the magnetization is in the equilibrium state, it does not produce a detectable MR signal.In order for a signal to be detected, the magnetization must first be disturbed from equilibrium.The energy needed to cause an energy transition or a change of the magnetic moments from being parallel to the field to anti-parallel is at the same frequency as the precession of the magnetic moments in that field.This is called the resonance condition.Protons precess at a resonance frequency that is proportional to B¡, and this is defined by the Larmor Equation: to6=yBewhere omega (o0) is the resonance frequency and gamma (y) is a gyromagnetic constant.Each type of nucleus has a specific spin and gyromagnetic ratio.By applying an oscillating magnetic field that stays in phase with the rotation of the magnetic moments and keeps pushing them in the same direction, this field is then also at the right energy to give energy to the magnetic moments.As a result, we can use a low intensity magnetic field, given that it is at the right frequency, to have a strong influence on the magnetization.At the equilibrium state, the magnetization is parallel to Be and does not move.To get it to move away from Be, a weak magnetic moment which is oscillating at the Larmor frequency is used.The Larmor frequency is in the radio frequency range, which is apploximately 64lMI{z at 1.5 Tesla.This magnetic field that is used to tip it away, 81, needs to be at a 90 degree angle to Bo.If it were parallel, it would only add to the field, but at a right angle it is able to cause the magnetization to precess out of alignment with Bo.The magnetization is precessing around the net field, and so by changing the direction of the field that we add at the same speed as the magnetization, this effect accumulates.The magnetization rotates away from the alignment with 86.With a bdef pulse of the B1 field it is possible to rotate the magnetization completely into the transverse (xy) plane.This brief pulse is called the radio-frequency (RF) pulse.The absorption of RF waves, which causes the spins to change their orientation from parallel to anti-parallel, is refened to as perturbation.With the application of an RF wave, M¡spirals down towards the transverse plane.\Vhen the RF is turned off, three simultaneous effects occur.One, the absorbed RF energy dissipates away into thermal energy as the magnetic moments relax back to equilibrium.The signal we detect is from the magnetic moments rotating in phase, to ploduce a time-varying magnetization that induces an electric cunent in an MR coil.Two, the excited spins begin to retum to odginal orientation (T1 relaxation).And th¡ee, the initially in-phase excited protons begin to dephase (T2 and T2* relaxation).Relaxation times are physical properties of the water environment (in terms of biological tissues).Relaxation refers to the retum of the spins to a stable low energy or random state after they have been excited or altered.Relaxation implies, therefore, that the spin system is retuming to a state of equilibrium.Further explanation of these concepts can be found in Bitar et al. (2006).There are a number of soutces of field variation over which the experimenter has no control.There is, however, a field variation intentionally produced by applying gradients.Gradients are produced by coils of wire situated in the magnet that can be tumed on and off.They are meant to produce a magnetic field that is parallel to Bo but vary linearly in magnitude along one of the axes.At the center of the magnet and at the center of the coil the magnetic field produced is zero.Moving along the axis, the field magnitude changes.Gradients can be produced in any direction by applying gradients in two or three directions at a time.With the gradients applied and an RF pulse applied on a particular frequency, it is possible to produce an effect on a selected region of space.This means that in one position in the field, the magnetic moments are precessing at whatever particular frequency the RF pulse is applied.On either side of this position rhe precessing frequency is different enough to no longer have an effect.If a gradient is applied in the y direction, moving away from the ze¡o center of the magnet toward the spot where the moments are precessing at the appropriate frequency (that of the RF pulse) creates an effect in a particular location.The bandwidth of the RF pulse will determine what range of frequencies will be affected.This is what enables slice selection.After the RF pulse has been applied, the moments have ¡otated away from the z axis to the transverse plane.However, b

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.