|
|
1
|
|
|
|
Jingfei Ma, PhD
University of Texas MD Anderson Cancer Center
Bracco Diagnostics/RSNA Research Scholar Grant
(2004 - 2006)
Differentiation between Benign and Malignant Vertebral Compression Fractures with Quantitative Diffusion and Fat MR Imaging
|
|
|
|
Abstract:
|
|
|
Benign and malignant vertebral compression fractures are both common in cancer patients with back pain. Early and correct determination of the lesions’ nature is especially important for clinical staging and therapeutic planning. Unfortunately, conventional magnetic resonance (MR) imaging techniques are sometimes inadequate for conclusively determining the lesions’ true etiology.
In recent years, diffusion-weighted MR imaging as well as gradient echo water and fat opposed-phase MR imaging (which probes relative fat content) have each separately been reported to be useful in characterizing vertebral lesions. The techniques used in the previous studies, however, are not quantitative and not optimal. T2 relaxation (in the case of diffusion-weighted imaging) and non-uniform water distribution (in the case of water and fat opposed-phase imaging) may complicate the imaging findings. Further, diffusion and fat are different parameters for a vertebral compression fracture and the extent of both may be altered by the presence or absence of metastasis, fracture edema and bone marrow. There has been no study comparing the value and difference of using diffusion and relative fat content for characterizing vertebral compression fractures.
We hypothesize that quantitative measurements of the apparent diffusion coefficient (ADC) and relative fat content can be used to better differentiate between benign and malignant vertebral compression fractures. To test this hypothesis, we have developed and optimized a quadratically phase-modulated diffusion-weighted single-shot fast spin echo imaging technique as well as a two-point spin echo Dixon water/fat imaging technique. The advantages of these techniques over those used in the previous studies include their capability for quantitative imaging, high signal-to-noise ratio and insensitivity to field homogeneity and motion. Our preliminary results indicate that these techniques are well suited for the proposed quantitative spine imaging. The specific aims of my project are as follows:
1. To measure the ADC of vertebral compression fractures of 48 cancer patients using the quadratically phase-modulated diffusion-weighted single-shot fast spin-echo imaging technique. We will determine whether the quantitative ADC value improves accuracy of lesion characterization over conventional MR imaging, using either clinical/imaging follow-up or pathology as the gold standard.
2. To measure in the same group of patients and during the same imaging session the relative fat content of the lesions relative to that of neighboring normal vertebral bodies using the two-point spin echo Dixon water/fat imaging technique. We will determine whether the quantitative fat content improves accuracy of lesion characterization over conventional MR imaging, using the same gold standard as is used for Specific Aim 1.
3. To determine the receiver operating characteristics (ROC) areas for quantitative ADC and fat content using the Mann-Whitney statistic and appropriate 95% confidence interval, and to compare the diagnostic differences between the two techniques using a two-sided paired t test.
Actively growing tumors are expected to contain little fat and to have lower diffusion than benign fracture edema due to differences in cellularity. A mixture of edema and bone marrow could also result in low diffusion but is expected to have substantial fat presence. Successful completion of the project will determine in the same group of patients the accuracy and the difference of the ADC and fat content in characterizing the nature of vertebral compression fractures. The results will be valuable in selecting an optimal technique and in potentially determining whether the diagnostic accuracy can further be improved with a combination of the techniques. Additionally, I expect that the techniques validated in this project will have utility for characterizing fractures in other anatomic regions.
|
|
|
More Activities by Jingfei Ma, PhD
|
|
|
|
2
|
|
|
|
Jingfei Ma, PhD
Johns Hopkins
RSNA Research Seed Grant
(1996 - 1997)
MRI Versus Bone Scan In the Evaluation of Unexplained Hip Pain
|
|
|
|
|
