Rapid Automated Corpus Callosal Magnetization Transfer Ratio and Morphometrics at 3T: Multiple Sclerosis Assessment

Joanna B. Jeong, MS1, Kenneth L. Weiss, M.D. 2, Maria J. Melanson, M.D.3, Mohammad Karim, BS1, Jun Ying, PhD3, Jane L. Weiss, BSN4

1 University of Cincinnati, College of Medicine

2 University of Cincinnati, Department of Radiology

3 University of Cincinnati, Department of Medicine

4 WestImage, Division of Research

We propose evaluating multiple sclerosis patients with a rapid computer automated MRI technique providing both corpus callosal magnetization transfer ratio (MTR) and morphometrics.

1.  Specific Aims

The long term goal is to promote patient care by providing a rapid reproducible imaging bio-marker of white matter involvement in multiple sclerosis (MS).

The central hypothesis is that controlling for age and gender, automated corpus callosal MTR and area measurements will correlate highly with MS duration, both negatively.

Specific Aim #1—assess MS involvement with MRI metrics: Disease duration will be compared to midline sagittal corpus callosal MTR and area measurements in 100 consecutive MS patients.

Specific Aim #2-- assess metrics' reproducibility and accuracy: To assure precision of measurements including stability of MRI system, metrics will be repeated in volunteers over time. Automated measurements will be compared to those derived by manual tracing of the corpus callosum for volunteers and MS patients alike.

2. Background and Significance

Three hundred thousand patients in North America suffer from multiple sclerosis, a debilitating demyelinating disease.(1) The corpus callosum, the largest white matter tract connecting the hemispheres, is significantly affected in multiple sclerosis, as evidenced by changes in volume and magnetization transfer. While the magnetization transfer ratio (MTR) has proved an important metric in the evaluation of MS; long acquisition times, SAR limitations at 3T, volume averaging, and time intensive post processing have limited routine clinical application.(2)

Dr.Weiss's research laboratory recently developed a rapid technique to provide automated callosal morphometrics and MTR at 3T, providing sub-millimeter in-plane resolution with total acquisition time less than 20 seconds. In a pilot study of three normal volunteers, highly reproducible hand-traced and computer automated corpus callosal area and MTR measurements were achieved from these rapid mid-sagittal acquisitions.(3) Numerous patients with multiple sclerosis have since received similar imaging sequencing, but related image processing and analysis have yet to be performed or correlated with clinical parameters.

It is our hope that this technique will facilitate assessment of patient's with MS, affording more optimized therapeutic intervention. This project will further my understanding of scientific research, advanced imaging techniques, and related post-processing; potentially jump-starting my future career in academic radiology.

3. Experimental Design

Methods & Data Collection

MRI shelf data from 100 patients followed in Dr. Melanson's MS clinic, who received sagittal MTR sequencing, will undergo retrospective image-processing and be analyzed in conjunction with clinical data to include patient age, gender, and disease duration. Additionally, the three normal subjects in the aforementioned pilot study will be re-imaged prospectively to further assess measurement reproducibility and scanner stability over time. Imaging data will include roll and yaw corrected, mid-sagittal SPGR sequencing , (TR 40 msec, TE 5.7-5.9 msec, flip angle 10, BW +/- 32 kHz, FOV =22, 256 x 224 matrix, 4 or 5 mm thick, 9 sec) with and without MT pulses acquired on a 3.0 T MRI with 8 channel head array coil ( GE, Waukesha, WI). MTR (Non MT - MT/ Non MT) images will be generated and segmented in Matlab (MathWorks, Natick, Mass) by applying a smoothing filter to the product of the MTR and inverted SSFSE images followed by a cluster algorithm. (Fig 1)

Analysis

Area and MTR measurements derived from auto-segmented (and hand callosal) tracings will be assessed of reproducibility using intra class correlation (ICC) and coefficient of variation (CV) respectively. Relationships between auto-segmented and hand callosal tracings will be assessed using correlation coefficients and/or multivariate regression models controlling for age and gender. Associations of corpus callosal area and MTR measurements to the MS duration will be assessed using multivariate linear models and/or multivariate fixed effect models.

References:

1. Wikipedia contributors. Multiple Sclerosis. In: Wikipedia, The Free Encyclopedia, 2010.

2. Inglese M, Horsfield MA, Filippi M. Scan-rescan variation of measures derived from brain magnetization transfer ratio histograms obtained in healthy volunteers by use of a semi-interleaved magnetization transfer sequence. AJNR Am J Neuroradiol 2001; 22:681-684.

3. Karim M, Weiss K, Ying J, Weiss J. Rapid automated corpus callosal MTR and morphometrics at 3T. In:ASNR. Boston, MA, 2010; submitted.

1