Comparison of the clinical standard of care (A) T1W image, (B) T2W image and (C) our proposed high-resolution sequence for a single subject. Comparison of the clinical standard of care (A) T1 image, (B) T2 image and (C) our proposed high-resolution sequence for a single subject.
Human Visual System Anatomy
- Eye globes
- Optic Nerve
- Optic Chiasm
- Lateral Geniculate Nucleus
- Diseases Affecting the Optic Nerve
The ON is a critical structure for vision and transports all visual information generated in the retina posteriorly through the optic chiasm. ON lesions anterior to the optic chiasm will impair vision in the ipsilateral eye.
Optical Coherence Tomography Imaging
Background and Techniques
The reflections from the two arms meet back at the beam splitter where they interfere. The path length of the reference arm can be swept in time with recordings of the interference at each reference arm length taken to obtain data at different depths.
Applications
RNFL thinning was found even in MS patients with a history of optic neuritis with good visual recovery in both affected and healthy eyes [ 30 , 31 ]. It is thought that RNFL becomes thinner in eyes without a history of optic neuritis due to neurodegenerative retinal ganglion cell degeneration.
Magnetic Resonance Imaging of the Optic Nerve
- MRI Background
- Challenges
- Applications
- Imaging History
The location of the ON directly above the maxillary sinuses causes significant sensitivity artifacts. This rapid motion also causes significant ON motion immediately behind the globe.
Magnetization Transfer Imaging
Magnetization Transfer Contrast
Modern methods have moved on to quantitative modeling of the exchange process to more accurately characterize the macromolecular content of the voxel. Reducing the MT phenomenon to a single measure, MTR, was known to be an oversimplification, and although MTR is a useful construct, even being used in clinical trials, by 2001 a quantitative formulation of MT exchange (qMT) had emerged [ 71].
Image Processing
Segmentation
Medical image segmentation has been applied to a wide variety of anatomy, but classical anatomy is of interest in brain segmentation. The brain was one of the first structures to be segmented due to the relative regularity between subjects, the rigid container of the skull, and traction applications.
Conjugate Gradient Descent
With all the atlas images registered in the target space, the segmentation from each atlas is propagated to the target through a process referred to as label fusion.
Random Forest Regression
Random forests are learned by training multiple decision trees on random subsets of the input features [85]. The accuracy of the forest can then be evaluated on test data that was not used to train the model and should correspond to the generalization error by applying the model to other yet unseen data.
Contributions
Previous Publications
A method for automatically measuring the radius of the ON and surrounding CSF by fitting the intensity model [87]. Using the previous two methods in healthy controls to develop a normative population useful for analyzing the patient population [88].
Robust Optic Nerve Segmentation on Clinically Acquired CT
- Introduction
- Methods
- Data Set
- Development Methods
- Evaluation Methods
- Results
- Development Results
- Inter-rater Reproducibility
- Evaluation Results
- Discussion
Quantities of automatic segmentations were calculated for ON and eyeball structure to identify deviations. Two subjects with tumors in the ON region caused oversegmentation in 17 of 33 failed scans.
Disambiguating the Optic Nerve and CSF Sheath
- Imaging
- Sequence Design
- Validation
- Analysis
- MRI Acquisition and Summary of Analysis Approach
- Proposed Model
- Initialization
- Optimization
- Calibration
- Validation
- Clinical Application
- Results
- Calibration
- Validation
- Clinical Application
- Discussion and Conclusions
The size of the optic nerve (ON) and cerebrospinal fluid (CSF) sheath was previously measured manually and was suggested as a differential diagnosis [142]. Balazsi, A., et al., The effect of age on the nerve fiber population of the human optic nerve. Cummings, T.J., et al., Gliomas of the optic nerve: histologic, immunohistochemical (MIB-1 and p53), and MRI analysis.
Optic Nerve and CSF Sheath Size Short-Term Reproducibility and Variability of Optic Nerve
Methodology
- Data
- Analysis
- Short-Term Reproducibility
- Demographic Analysis
The radius of the ON (bottom line) and cerebrospinal sheath (top line) can be seen in Figure IV.2 for 8 subjects in scan 1 (blue) and 8 subjects in scan 2 (red). In Figure IV.4, the ON of patient 1 appears to be smaller than the normative distribution around the midpoint of the ON. The right eye remains atrophic throughout ON, whereas the left eye appears to approach the normative distribution.
Improved Automatic Optic Nerve Radius Estimation from High Resolution MRI
- Proposed Method
- Data Acquisition
- Analysis: Short- and Long-Term Reproducibility
- Long-Term Reproducibility
Weigel, M., et al., Fast and quantitative high-resolution magnetic resonance imaging of the optic nerve at 3.0 Tesla. Gerling, J., et al., Optic nerve diameter in idiopathic optic neuritis and in anterior ischemic optic neuropathy. Geeraerts, T., et al., Use of T2-weighted magnetic resonance imaging of the optic nerve sheath to detect increased intracranial pressure.
Quantitative Characterization of Optic Nerve Atrophy in Patients with Multiple Sclerosis
Materials and Methods
- Study Design
- MRI Protocol
- Volumetric Measurements
- Statistical Analysis
- Qualitative Results
- Quantitative Analysis
Van Leemput, K., et al., Automatic model-based tissue classification from MR images of the brain. Bekes, G., et al., Geometric model-based segmentation of the organs of vision on CT images. Shen, S., et al., Normative measurements of the Chinese extraocular musculature by high-field magnetic resonance imaging.
Characterization of Short- and Long-Term Stability of Non-Invasive Optic Nerve Imaging 84
MRI Results
Raw short (left) and long term. right) reproducibility for four selected metrics showing T2w (red) and T1w (blue) measurements to illustrate the reliability and stability of different metrics, as well as T1w/T2w bias (bottom right). Reproducibility as percentage difference between corresponding short-term and long-term reproducibility scans for T1w and T2w scans for the mean diameter of the eight rectus muscles. The short- and long-term percentage differences were not significantly different for T1w and T2w measurements (Wilcoxon rank sum; p<0.01).
OCT Results
Discussion and Conclusions
Henderson, A.P., et al., A study of the retinal nerve fiber layer in progressive multiple sclerosis using optical coherence tomography. Steinborn, M., et al., High-resolution ultrasound and magnetic resonance imaging of the optic nerve and optic nerve sheath: anatomical correlation and clinical importance. Hajnal, JV, et al., Use of fluid-attenuated inversion recovery (FLAIR) pulse sequences in brain MRI.
Structural-Functional Relationships between Clinical Eye Orbital MRI and Visual
Patient Data
The relationship between visual function and MRI-derived orbital structures was investigated in a retrospective cohort of patients at Vanderbilt University Medical Center. Each of the sessions had an example of visual function testing available within 6 months of the scan. Visual function was assessed using the American Medical Association Functional Vision Score (FVS), which is calculated based on visual acuity and visual field testing obtained through routine clinical care.
Image Processing
Statistical Modeling
Image Processing
Statistical Modeling
Discussion and Conclusion
Hickman, S.J., et al., A serial MRI study following the median area of the optic nerve in acute optic neuritis. Harrigan, R.L., et al., Characterization of short- and long-term stability of noninvasive optic nerve imaging.
Numerical Optimization of Quantitative Magnetization Transfer Sampling Schemes
Background
For simplicity, we assume that the direct effect of the off-resonance saturation on the free pool is negligible and the order is . Using Equations (IX.8), (IX.13) and (IX.16) we can solve for the z-component of the free-pool magnetization directly before the readout pulse which will be proportional to our signal intensity. Using these terms, we can write the z component of the free pool magnetization before the
Methods
- qMT Model Implementation
- Optimization
- Acquisition
- Processing
Once all 16 points in the original model had been evaluated for superior performance, the final model consisted of the sampling scheme in Table IX-1, column 2. Gray matter (GM) and white matter (WM) were manually labeled on the mFFE scan and propagated to each of the qMT fits for comparison of PSR contrast. The optimized sequence shows lower variance within WM than the other two methods, while maintaining the superior contrast of the high-displacement approach.
Discussion
Some acquisition schemes can be very robust in distinguishing tissue types, but only if they are acquired accurately. However, this could be extended to include more RF power, assuming imaging parameters such as RF deposition and TR are still reasonable. Also, each of the RF powers can be more densely sampled to include more than 32 points.
Conclusions and Future Work
Summary
Multi-Atlas Optic Nerve Segmentation
- Summary
- Main Contributions
- Future Work
Multi-atlas segmentation relies on matching contrast between atlas images and target images, which is not true for different contrast mechanisms in MRI. We extend this algorithm to automatically calculate both radiological and morphological metrics describing the eye's orbital structures. We use the volumetric labels produced by the multi-atlas segmentation technique to automatically calculate various morphological features of each eye's orbital structure, including radiologically defined biomarkers of disease progression.
Non-Invasive Automatic Optic Nerve Radius Estimation
- Summary
- Main Contributions
- Future Work
We proposed and optimized the first automatic ON radio estimation algorithm that exploits the superior contrast from this optimized pulse sequence. We propose a method to compare these ON measurements by interpolating the measurements to be the same length, by aligning parts of the ON, as is commonly done in other anatomies such as the spinal cord. Improving the accuracy and robustness of the radius estimation algorithm by further improving the understanding of the manifold between model parameters and radius space could improve accuracy and clinical utility.
Short- and Long-Term Optic Nerve Imaging Reproducibility
- Summary
- Main Contributions
- Future Work
With the release of these data, the host of a challenge would now be possible to push the community further towards developing clinically relevant imaging techniques. The inclusion of the current clinical standard of care imaging allows for the evaluation of imaging techniques that can be used to take advantage of the large volumes of clinical imaging currently stored in medical centers around the world. In addition, the current state-of-the-art imaging included in this study allows investigators without the ability to acquire their own sophisticated ocular circuit MRI to develop imaging techniques and freely evaluate them.
Numerical Optimization of qMT Sampling
- Summary
- Main Contributions
- Future Work
The publication of normative values makes a positive contribution to the literature on noninvasive imaging of the ocular orbit, allowing comparison of the techniques and measurements of other researchers. We show that this optimized lag sampling scheme displays superior homogeneity of pool size ratio estimation within white matter and gray matter and superior contrast between white matter and gray matter in healthy control spinal cord compared to traditional qMT sampling schemes. While this framework initially presents promising results for improving qMT sampling schemes, there are many areas that need further investigation.
Concluding Remarks
Journal Articles
Conference Publications
Structural-Functional Relationships Between Ocular Circuit Imaging Biomarkers and Clinical Visual Assessments” In Proceedings of the SPIE Medical Imaging Conference. 34;One the Fallacy of Quantitative Segmentation for T1-Weighted MRI." In Proceedings of the SPIE Medical Imaging Conference. 34;Construction a Statistical Atlas of the Radii of the Optic Nerve and Cerebrospinal Fluid Sheath in Young Healthy Adults." In conjunction with the SPIE Medical Imaging Conference.
Abstracts
Tripp, S.A., et al., Axonal loss of the retinal nerve fiber layer and visual dysfunction in optic neuritis.