Date of Award

Summer 2000

Document Type

Thesis - Restricted

Degree Name

Master of Science (MS)


Biomedical Engineering

First Advisor

Ropella, Kristina M.

Second Advisor

Donahue, Kathleen M.

Third Advisor

Ulmer, John L.


PURPOSE: The goal of this preliminary study was to quantitatively compare FLAIR DWI and "conventional" DWI for clinical utility in stroke diagnosis and evaluation. METHODS: We retrospectively analyzed data from 18 stroke patients that were imaged 14 to 103 hours after symptom onset. Conventional and FLAIR prepared isotropic DW spin echo (SE) echo planar (EP) images with gradient factors of b=O and b=lOOO s/mm2 were examined. We used three-dimensional (3-D) semi-automated detection techniques to identify lesions and statistically quantify and compare the signal intensity ratio (SIR), contrast-to-noise ratio (CNR), apparent diffusion coefficient (ADC), and ADC ratio (ADCr) for both techniques. RESULTS: The SIR and CNR were significantly higher for Conventional DWI than for FLAIR DWI. Lesion ADCs were not significantly different for the 2 techniques; however, ADCs from normal contralateral ROls were significantly different. This led to significant differences in the ADCr values. In addition, FLAIR DW images correlated better to ADCs than did conventional DW images. CONCLUSION: Our study indicates that lesion conspicuity is better with conventional DWI while actual ADC "accuracy" may be better with FLAIR DWI. This would suggest that initial examination and diagnosis should use conventional DWI, but FLAIR DWI may offer advantages for more detailed quantification of diffusion characteristics (e.g. evaluation of neuroprotective agents and thrombolytic therapies).



Restricted Access Item

Having trouble?