Comparison of singular value decomposition and Fourier deconvolution methods forcerebral blood flow quantification in dynamic contrast-enhanced magnetic resonanceimaging

Introduction: glioma grading is an important clinicalprocedure that determines the procedure of the patient'streatment. there is an increasing interest in othercomplementary techniques in addition to histo-pathologicalevaluation as the gold standard method. perfusion indices haveshown promising correlations with histo-pathological gradesand neovascularization degrees in gliomas. in this study, theprocessing times and classification performances of cerebralblood flow (cbf) parameters calculated using singular valuedecomposition (svd) and fourier deconvolution methodswere evaluated.materials and methods: the statistical differences betweenthe cbf magnitudes of the high- and low-grade gliomas wereevaluated for eighteen patients with pathologically provengliomas (low-grade gliomas, 5; high-grade gliomas, 13). theclassification performances of the calculated indices wereevaluated using the receiver operating curve analysis.results: the deconvolution processing times for the svdmethod were significantly higher than those of the fourierdeconvolution method (about ~9 times). there werestatistically significant differences in both cbf indicesbetween the high- and low-grade gliomas (p<0.05). in roccurve analysis, cbf indices calculated by the svd method hada higher area under the curve (about 8538.5 vs. 8153.8).conclusion: based on the study, the glioma grade might bedetermined using the fourier-based cbf index with an almostsimilar grading efficiency and a much lower processing timethan the cbf index calculated by the svd method.