Retinal nerve fiber layer is associated with brain atrophy in multiple sclerosis

OBJECTIVE

Optical coherence tomography (OCT) noninvasively quantifies retinal nerve fiber layer (RNFL) thickness. Studies show RNFL thinning in multiple sclerosis (MS), and we assessed its association with brain atrophy.

METHODS

RNFL thickness was measured in 40 patients with MS and 15 controls. Brain parenchymal fraction (BPF) and partial brain volumes were estimated from cranial MRI scans using SIENA-X. Multiple linear regression modeling assessed the association between OCT and MRI measures of atrophy.

RESULTS

Minimum RNFL thickness and subject age together predict 21% (p = 0.005) of the variance in BPF in all patients with MS and 43% (p = 0.003) of the variance in BPF in the subgroup with relapsing remitting MS (RRMS; n = 20). The partial correlation coefficient between BPF and minimum RNFL thickness, controlling for age, is 0.46 (p = 0.003) in all patients with MS and 0.69 (p = 0.001) in patients with RRMS. These associations are driven by CSF volume but not by gray or white matter volume. There is no significant association of these variables among controls.

CONCLUSIONS

In multiple sclerosis (MS), retinal nerve fiber layer thickness is associated with brain parenchymal fraction and CSF volume. These data suggest that quantification of axonal thickness in the retina by optical coherence tomography (OCT) provides concurrent information about MRI brain abnormality in MS. OCT should be examined in longitudinal studies to determine if it could be used as an outcome measure in clinical trials of neuroprotective drugs.

Precision and accuracy of regional radioactivity quantitation using the maximum likelihood EM reconstruction algorithm

The imaging characteristics of maximum likelihood (ML) reconstruction using the EM algorithm for emission tomography have been extensively evaluated. There has been less study of the precision and accuracy of ML estimates of regional radioactivity concentration. The authors developed a realistic brain slice simulation by segmenting a normal subject's MRI scan into gray matter, white matter, and CSF and produced PET sinogram data with a model that included detector resolution and efficiencies, attenuation, scatter, and randoms. Noisy realizations at different count levels were created, and ML and filtered backprojection (FBP) reconstructions were performed. The bias and variability of ROI values were determined. In addition, the effects of ML pixel size, image smoothing and region size reduction were assessed. Hit estimates at 3,000 iterations (0.6 sec per iteration on a parallel computer) for 1-cm(2) gray matter ROIs showed negative biases of 6%+/-2% which can be reduced to 0%+/-3% by removing the outer 1-mm rim of each ROI. FBP applied to the full-size ROIs had 15%+/-4% negative bias with 50% less noise than hit. Shrinking the FBP regions provided partial bias compensation with noise increases to levels similar to ML. Smoothing of ML images produced biases comparable to FBP with slightly less noise. Because of its heavy computational requirements, the ML algorithm will be most useful for applications in which achieving minimum bias is important.