Renal artery stenosis: extracting quantitative parameters with a mathematical model fitted to magnetic resonance blood flow data.

PURPOSE To investigate the feasibility of quantitative parameter extraction from a mathematical model fitted to renal artery magnetic resonance flow data. MATERIAL AND METHODS A total of 16 subjects, eight patients, and eight normal controls, were examined with cine phase-contrast velocity measurements, and blood flow data from the aorta and both renal arteries were extracted by means of contour detection. A mathematical model with eight parameters describing the time, duration, and amplitude of the systolic acceleration and the diastolic deceleration was fitted to the aorta and renal artery blood flow data from each subject. The curve fitting was evaluated with R(2) values. Statistical analysis was performed with unpaired Wilcoxon tests and stepwise logistic regression. RESULTS A total of three data sets out of 48 yielded R(2) values below 0.80 and were considered unreliable for parameter estimation. Basal flow was significantly, and systolic peak amplitude almost significantly, lower in stenotic arteries. Logistic regression indicated that two parameters describing basal flow and the duration of acceleration can accurately predict stenosis. CONCLUSION The results suggest that it is technically feasible to fit a mathematical model to renal blood flow data, extracting quantitative parameters that may prove useful for quantification and diagnosis of renal artery stenosis.