The relations between brain areas involved in vision were explored in 8 patients with unilateral acute optic neuritis using functional magnetic resonance imaging (fMRI) and diffusion tensor imaging (DTI). In all patients monocular stimulation of affected and unaffected eye elicited significantly different activation foci in the primary visual cortex (V1), whereas the foci evoked in the middle temporal visual area (area V5) were similar in size and in delay of blood-oxygen-level-dependent response. DTI analysis documented lower white matter anisotropy values and reduced fibre reconstruction in the affected compared with the unaffected optic nerves. The preserved activation of area V5 observed in all our patients is an interesting finding that suggests the notion of a different sensitivity of the optic pathways to inflammatory changes.
Optic neuritis (ON) is an inflammatory disease of the optic nerves characterized by retrobulbar pain, unilateral vision impairment or loss, visual field defects, and impaired contrast and colour sensitivity [
fMRI and diffusion tensor imaging (DTI) data were collected in 8 patients with acute ON (mean age 46 years, 5 women; Table
Characteristics of the optic neuritis patients.
Patient | Age | Gender | Affected eye | Underlying pathology |
---|---|---|---|---|
(years) | ||||
1 | 70 | F | Left | Idiopathic ON |
2 | 50 | F | Left | Multiple sclerosis |
3 | 45 | M | Left | Facial trauma |
4 | 37 | M | Left | Multiple sclerosis |
5 | 23 | F | Right | Idiopathic ON |
6 | 37 | M | Right | Multiple sclerosis |
7 | 81 | F | Right | Ischemia |
8 | 28 | F | Right | Multiple sclerosis |
Exhaustive ophthalmological examination and perimetry testing demonstrated unilateral vision loss or impairment (4 in the left eye and 4 in the right eye). Visual acuity in the affected eye ranged from 20/200 (1.0 LogMAR) to 20/125 (0.8 LogMAR) in all patients.
MR coronal images obtained with STIR sequences showed hyperintense areas in the affected optic nerves in all patients (Figure
Coronal images of the orbits of two patients with acute right (a) and left (b) optic neuritis obtained with an inversion recovery (STIR) sequence. (a) The anterior part of the right optic nerve is swollen. The perioptic space is less obvious. (b) The posterior portion of the left optic nerve is bright and swollen. Yellow arrows: affected optic nerve. L and R: left and right.
Subjects were placed in a 1.5-Tesla (T) scanner (Signa Excite NV/i CV/i, General Electric Medical System, Milwaukee, WI, USA) equipped with 50 mT/m gradients with their head restrained within a circularly polarized head coil. Morphological brain MR images were obtained using sagittal Spin-echo T1-weighted sequences, axial FLAIR and T2-weighted Fast Spin-echo sequences, and coronal T2-weighted sequences. The optic nerve was explored using a coronal STIR sequence (section thickness, 3 mm).
The experimental procedure consisted in the acquisition of a 3D data set (IR Prep Fast SPGR 3-D; TR 15.2 ms, TE 6.9 ms, TI 500 ms, Flip Angle 15°, FOV 29 × 29 cm, slice thickness 1 mm, matrix 288 × 288, 1 Nex, and scan time 8:20 min) and of 10 contiguous 5-mm-thick axial high-resolution anatomical images (T1 FLAIR, 2D, TR 1700 ms, TE 24 ms, Field of View 24 × 24 cm, thickness 5 mm, Matrix 256 × 256, 1 Nex, scan time 2:25 min for 10 images) on which the functional activations were overlaid. fMRI data were then acquired in the same axial planes with a single-shot T2*-weighted gradient-echo EPI sequence (TR 3000 ms, TE 60 ms, Flip Angle 90°, Field of View 24 × 24 cm, Matrix 64 × 64, 1 Nex, scan time 5:12 min) to obtain, during the stimulation cycle, 1000 axial functional images (100/section, 1 image/3 s) from the 10 contiguous 5-mm-thick axial sections selected.
Axial diffusion-weighted images with a diffusion sensitising gradient (
Visual stimuli-generated using an in-house software were projected in a head-mounted display. A black and white chessboard (amplitude: 6°) was presented to the centre of the visual field (CVF) to an eye at a time (virtual Cartesian distance from viewer’s eyes: 75 cm) using a 5 min block-design experimental paradigm alternating twenty 15-s periods of rest and stimulation. Patients were instructed to fix their gaze on a cross in the centre of the display throughout the visual studies. Eye movements were monitored with an internal camera.
After each experimental session the images were transferred to a Unix workstation (General Electric Advantage Windows 4.2) and then to a personal computer. Data were analyzed with the BrainVoyager QX software (Brain Innovation, Maastricht, the Netherlands). The first two images of each functional series were discarded to take into account the period of signal intensity variation related to progressive saturation. Data from each subject were preprocessed to remove noise and artefacts. The functional images of each subject were overlaid on the 2D anatomical images and coregistered into their 3D data sets through trilinear interpolation. Data were then transformed into Talairach space [
Statistical analysis was applied to the data from each patient using the general linear model (GLM) [
The images were transferred to the Unix workstation, where DTI data were postprocessed using Functool 3.1.22 (GE Medical Systems). Echo-planar imaging distortion was automatically corrected. Diffusion eigenvectors and eigenvalues, reflecting the main direction of diffusion and associated diffusivity, were calculated from the diffusion tensor. Anisotropy was calculated using orientation-independent fractional anisotropy (FA). The FiberTrak option allows Functool DTI processing to create 2D colour orientation maps, 2D colour eigenvector maps, and 3D tractography maps. The 3D volume viewer permits areas with high FA to be displayed as 3D images. The anisotropy threshold for tracking termination was 0.18. Regions of interest (ROIs) measuring 105 mm2 were selected in the occipital visual cortex to visualize the optic radiations. Other ROIs were selected in the retrobulbar segment of the optic nerve and along the retrochiasmatic pathway, to reconstruct the fibres crossing through them (using the origin and destination regions method). All ROIs were defined manually on axial colour-coded maps of the main diffusion directions.
The mean volume of the activation foci and the mean signal increase elicited by stimulation of affected and unaffected eyes was calculated and subjected to a paired
Visual stimulation of the unaffected eye evoked activation foci in occipitotemporal and posterior parietal areas in both hemispheres (Figure
Activation foci evoked by stimulation of the unaffected (a) and affected eye (b) in a patient with acute left optic neuritis. Stimulation of the unaffected side induced bilateral activation of the visual cortex; stimulation of the affected side was associated with a reduced voxel size and BOLD signal in the activation foci. L and R: left and right.
Activation foci evoked by stimulation of the unaffected (a) and affected (b) eye in a patient with acute left optic neuritis. White lines show activation in area MT/V5 in right and left hemispheres. The Talairach coordinates (
The volume size of the foci elicited in area V1 of both hemispheres, that is, the number of voxels activated during stimulation, ranged from 345 to 879 (mean, 576) in the unaffected eye and from 0 to 360 (mean, 70) in the affected eye; the difference was significant (
The BOLD signal increase evoked in area V1 by stimulation of the unaffected eye ranged from 0.9 to 1.7% (mean, 1.32%) and was significantly greater (
Voxel number (and percent BOLD signal increase) of the activation foci elicited in areas V1 and V5 during stimulation of the affected and unaffected eye.
Patient | ON side | Area V1 | Area V5 | ||
---|---|---|---|---|---|
Left side stimulation | Right side stimulation | Left side stimulation | Right side stimulation | ||
1 | L | 33 (1.65) | 345 (1.7) | 322 (0.7) | 365 (1) |
2 | L | 360 (1.6) | 680 (1.7) | 385 (1.33) | 338 (1.25) |
3 | L | no | 515 (1.5) | 700 (0.85) | 650 (0.9) |
4* | L | no | yes (—) | yes (—) | yes (—) |
5 | R | 530 (1.5) | 34 (1.5) | 726 (1) | 614 (1) |
6* | R | yes (—) | no | yes (—) | yes (—) |
7 | R | 509 (0.9) | 36 (0.6) | 407 (0.95) | 424 (0.9) |
8 | R | 879 (1.3) | 64 (0.6) | 493 (0.67) | 450 (0.7) |
Data regard activation foci contralateral to the stimulation side.
*For technical reason only visual localization and evaluation of activation foci was possible in patients 4 and 6.
L: left; R: right.
Number of activated voxels in areas V1 (a) and V5 (b) following stimulation of the affected (grey bars) and unaffected (white bars) eye in 6 patients (data of patients 4 and 6 are not provided due to bad quality of the 3D scans).
When the same measures were calculated in the foci elicited in area V5, the differences in volume size and the delay of the BOLD response evoked by visual stimulation of the unaffected and affected eye were not significant (resp.,
In this study DTI was applied as an adjunct to conventional MRI [
Mean FA values of optic radiations on the affected (grey bars) and unaffected side of all optic neuritis patients. Partial overlapping of error bars indicate a nonsignificant difference between affected and unaffected side anisotropy.
Optic radiations (yellow arrows) reconstructed using DTI fibre tracking in a patient with left optic neuritis. The two fibre bundles arise from the visual cortex and mingle with the inferior longitudinal fasciculus and inferior fronto-occipital fasciculus. Fibres also cross the splenium of the corpus callosum (cc). ot: optic tracts. L and R: left and right.
Optic nerve DTI in patients with acute right (a) and left (b) optic neuritis. DTI was able to show optic fibres in the unaffected optic nerve. At the site of the optic nerve lesion fibres cannot be reconstructed due to FA alterations; in some cases this helped define more precisely the site of the optic nerve lesion responsible for the visual impairment. L and R: left and right.
fMRI can be used to document functional damage and recovery in patients with optic nerve damage [
Area V5 plays an important role in motion perception and in the integration of local signals into a global perception; it is also involved in the guidance of some eye movements and projects to eye movement-related areas in the frontal (FEF) and parietal lobes (lateral intra-parietal area) [
The present study describes the cortical responses elicited in patients with a damaged optic nerve but unaffected retrochiasmatic pathways and a healthy visual cortex.
The two main pathways involved in vision [
The different functional activation detected in areas V1 and V5 in our patients may be related to differences in the functional specialization and physical characteristics of the relevant pathways, the ventral and the dorsal stream. Since the dominant input to area V5 is from the magnocellular stream [
The authors are grateful to Dr. Silvia Modena for language revision.