Renal Artery Doppler Ultrasound Diagnosis Research

ObjectionThe aim of this study was to certify the accuracy of SVM classfier of renal artery stenosis spectrum.Materials and MethodsFrom August2011to April2013,40patients with80renal arteries were detected by color Doppler sonography(CDS) and were referred to renal digital-subtraction angiography(DSA) or computed tomography angiography(CTA) afterwards. Six Doppler parameters, including renal peak systolic velocity(RPSV), renal-aortic ratio(RAR), renal-interlobar ratio(RIR), acceleration time(AT),acceleration(AC), resistance index(RI) were measured. At the same time the blood flow signal curve of interlobar artery was collected and stored. Then80spectrums of renal artery were used to testify the validation of support vector machines(S VM) classifier, and statistical analysis to determine the best threshold of the six parameters for predicting RAS was performed with receiver operating characteristic(ROC) curves. The sensitivity, specificity, negative predicting value, positive predicting value and accuracy of different parameters were calculated. The agreement between SVM classifier and the best parameters was measured using Kappa statistics.Results1. In the80main renal arteries demonstrated by renal arteriography, there were37cases(diameter reduction>70%),43cases (diameter reduction<70%or normal). ROC analysis showed that the AT and AC were the best parameters with the accuracy of95%, the sensitivity and specificity of AT were97.3%and93%respectively, that of AC were94.6%and95.3%respectively. The best cutoff values for the6parameters (RPSV, RAR, RIR, AT, AC, RI) were277cm/s,2.75,9.3,0.074s,165cm/s2,0.59respectively.2. Using RBF kernel, SVM classifier with five-dimensional features (curve first moment, second moment, peak H, fluctuations in the area of accounting for R, systolic quadratic fit parameters) as classification feature could effectively classify spectrum of RAS with94.6%sensitivity,93.0%specificity and93.8%accuracy.3. The agreement between AC and AT, AC and SVM, AT and SVM were good, their Kappa were0.95,0.88,0.83respectively.ConclusionsFor the detection of RAS (diameter reduction>70%), AT and AC are the best indicators in the six parameters. SVM classifier can effectively classify spectrum of renal artery and its diagnostic efficiency is close to the best indicators(AT and AC). The development of SVM classifier is a meaningful exploration in computer-aided diagnosis of RAS, which will help to promote the use of Doppler ultrasound for detecting RAS in the primiary hospitals. ObjectiveThe aim of this study was to evaluate the influence of posture and normal respiration on the renal artery blood flow parameters and provide the evidence for standardization of renal artery Doppler ultrasound.Materials and MethodsSex-matched healthy young people of60cases, respectively, in the supine and lateral position, both main renal artery and interlobar artery were dectected by Doppler ultrasound. During end of inspiration breath-holding and end of expiration breath-holding, peak systolic velocity(PSV), end-diastolic velocity(EDV), resistance index(RI) of bilateral main renal artery and PSV, EDV, RI, acceleration time(AT), acceleration(AC) of bilateral interlobar artery were measured and the parameters were analyzed. Doppler spectra of main renal artery were obtained at the origin of artery and Doppler spectra of interlobar artery were obtained at the upper, middle, and lower pole interlobar arteries. The renal artery blood flow parameters were also analyzed between both sides.Results1. PSV and RI of main renal artery between supine position and lateral position had statistically significant difference(p<0.001) and θ angle between supine position and lateral position also had statistically significant difference(p<0.001). EDV of main renal artery and PSV, EDV, RI, AC, AT of interlobar artery between supine position and lateral position had no significant difference.2. PSV, EDV, RI of main renal artery and PSV, EDV, RI, AC, AT of interlobar artery between left and right side had no statistically significant difference in supine position. In lateral position, PSV of main renal artery and their θ angle between right and left side had statistically significant difference(PSV:p=0.012,0:p<0.001), while the other parameters between two sides had no statistically significant differences.3. During the normal respiration, the positon of kidney changed slightly in vertical direction, but the renal artery had no obvious change. PSV, EDV, RI of bilateral main renal artery and PSV, EDV, RI, AT, AC of bilateral interlobar artery measured during end of inspiration breath-holding and end of expiration breath-holding had no significant difference.Conclusions1. PSV and RI of main renal artery between supine and lateral position had statistically significant difference. This showed that the postural factors affected the renal artery blood flow parameters and it should be considered when we checking renal artery with ultrasound. But it had no effect on the renal interlobar artery blood flow parameters.2. Except PSV of main renal artery in lateral position, the other parameters of main renal artery and interlobar artery between right and left side had no statistically significant differences.3. The renal artery blood flow parameters had no significant difference between different breathing phases in normal respiration. This results suggests that the effect of breathing in normal respiration could be neglected in Doppler measurement of the renal artery.