Unenhanced MR Angiography Imaging Techniques & Computational Fluid Dynamic Analysis Of Renal Artery

Part OneUnenhanced MR Angiography of Renal Artery in Hypertensive Patients Compared with Contrast-enhanced MR AngiographyObjective This study aims to evaluate the accuracy of unenhanced respiratory-gated TrueFISP ("True Fast Imaging with Steady State Precession") magnetic resonance angiography (MRA) and compare with contrast-enhanced MRA (CE-MRA) for diagnosis of renal artery stenosis (RAS) in patients with suspected secondary hypertension.Materials and MethodsForty-four patients (32 male,12 female) with hypertension and suspected RAS underwent both unenhanced and enhanced MRA within the same session on a 1.5-T MRI system for assessing renal arteries. All images were processed using medical visualization software and commercial processing package. The volume of main renal arteries, length of maximal visible renal arteries, number of visualized branch vessels, stenosis grade, and subjective quality were compared. The paired two-tailed Student t test was applied to evaluate the significance for continuous variables. The Wilcoxon signed rank test was used to analyze the ordinal scaled image quality scores. TrueFISP MRA and CE-MRA images were evaluated by two experienced radiologists using a joint reading performed in consensus.ResultsEighty-eight paired renal artery images were successfully acquired from TrueFISP MRA and CE-MRA techniques in all 44 patients. The volume of main renal arteries, length of maximal visible renal arteries, and number of renal artery branches indicated no significant difference (P=.62, P=.59, P=.91, respectively). For assessment of the severity of stenosis, CE-MRA indicated 26 RAS of total 88 renal arteries with stenotic grades from mild to severe and TrueFISP MRA showed only 20 RAS with stenotic grades from mild to moderate. And six false-positive results were demonstrated in the images from CE-MRA. The stenotic degree of six segments was concordant between two techniques. And in 20 segments with no consistency, the perceived degree of stenoses was greater in CE-MRA images than that in TrueFISP MRA. There was no significant difference of the image quality of proximal segments between the two techniques (P>.05). Qualitative scores for VR and MIP images, middle and distal segments of the main renal arteries, and their first branches, were higher on images from TrueFISP MRA than those from CE-MRA (P<.05).ConclusionUnenhanced TrueFISP MRA is a reliable method for evaluating renal artery disease. TrueFISP MRA shows an equal and even better image quality as CE-MRA and can be used as an alternative to CE-MRA for screening renal artery, especially in hypertensive patients with renal insufficiency.The combination of unenhanced MRA with CFD is an attempt to utilize completely noninvasive methods for comprehensive evaluation of renal artery stenosis. Our initial application of CFD on renal artery demonstrates its potential for an improved understanding of the link that exists between hemodynamics and renal artery stenosis. It may provide useful information regarding patient stratification and strategy in stenotic grades from mild to severe and TrueFISP MRA showed only 20 RAS with stenotic grades from mild to moderate. And six false-positive results were demonstrated in the images from CE-MRA. The stenotic degree of six segments was concordant between two techniques. And in 20 segments with no consistency, the perceived degree of stenoses was greater in CE-MRA images than that in TrueFISP MRA. There was no significant difference of the image quality of proximal segments between the two techniques (P>.05). Qualitative scores for VR and MIP images, middle and distal segments of the main renal arteries, and their first branches, were higher on images from TrueFISP MRA than those from CE-MRA (P<.05).ConclusionUnenhanced TrueFISP MRA is a reliable method for evaluating renal artery disease. TrueFISP MRA shows an equal and even better image quality as CE-MRA and can be used as an alternative to CE-MRA for screening renal artery, especially in hypertensive patients with renal insufficiency.The combination of unenhanced MRA with CFD is an attempt to utilize completely noninvasive methods for comprehensive evaluation of renal artery stenosis. Our initial application of CFD on renal artery demonstrates its potential for an improved understanding of the link that exists between hemodynamics and renal artery stenosis. It may provide useful information regarding patient stratification and strategy inPart TwoHemodynamic Analysis of Renal Artery Stenosis Using Computational Fluid Dynamics Technology Based on Unenhanced Magnetic Resonance AngiographyObjective Patient-specific computational fluid dynamics (CFD) technology, based on either MR or CT images, is one of the novel technologies to quantitatively estimate hemodynamic conditions. Although it has been successfully applied in different aspects of cardiovascular system, there has been no report to determine the hemodynamic conditions of in vivo renal artery stenosis (RAS) with this technology. This study aims to evaluate the feasibility of computational fluid dynamics technology in analysis of RAS based on unenhanced MR angiography (MRA).Materials and Methods1、Thirty consecutive patients (mean systemic blood pressure 160.8±22.6 mmHg) diagnosed with unilateral RAS, free from elevated serum creatinine and decreased estimated glomerular filtration rate, were prospectively enrolled in this study. Also 10 normal volunteers were recruited, scanned and processed using the same methods.2、All patients and volunteers underwent MRA examinations with a 1.5T MR imager (MAGNETOM Aera, software version syngo MR D11; Siemens Healthcare, Erlangen, Germany) equipped with 2 body coils placed both at the front (18-element body matrix) and back (12-element spine matrix) of the abdomen. Respiratory-gated unenhanced MRA (steady-state free precession, SSFP; Siemens Medical Solution) was performed. The acquisition time ranged from 5 to 7 min.3、Twelve of 30 patients underwent US (IU22, Philips Medical Systems, Netherlands). Color Doppler flow examination was performed by an ultrasonographer and peak systolic velocity (PSV) of stenotic main renal artery was obtained under resting conditions in 3 days after MRA examination.4、CFD TechnologyA commercial segmentation software Mimics (Version 14.0, Materialise Company, Leuven, Belgium) was utilized to transfer MRA images (DICOM format) into vessel-shaped three-dimensional geometric data (STL format). The STL artery surface data was meshed using ICEM (ANSYS, Lebanon, New Hampshire) to form grids for CFD simulation. Three layers of growth mesh (prism mesh) were meshed at gradient distances, gradually moving them away from the vessel to handle the boundary layer efficiently. The distance of the first layer to the vessel surface was fixed at 0.02 mm. The number of elements in all patient-specific models ranged from 2060198 to 3608948. The process of segmentation and mesh generation took 5-10 minutes.Flow entry and exit from the calculation domain were assumed to be normal to the inlet and outlet surface regions of the arterial geometry. To form a fully developed flow boundary layer at the proximal inlet, the inlet of the domain in an upstream direction was extended to 15 times the inlet diameter size so that fully developed velocities profiles were formed in the boundary layer. To simulate the peripheral capacitance, all outlets at the distal ends in the normal downstream direction were extruded to 40 times the vessel diameter size, sufficient for the recovery of blood pressure. These boundary conditions ensured that the length of the extended domain would not further influence the flow in the main-body domain. The extended domains at the inlet and outlets were only for the purpose of the simulation and were not included in the analysis of flow characteristics.The vessel wall was modeled as rigid and the assumption of zero wall motion was utilized in calculations. The fluid in this study was assumed as an incompressible Newtonian fluid, while the flow was assumed to be laminar. The density was set 1060 kg/m3, and the dynamic viscosity was 0.0035 Pa.s. Steady state flow condition was introduced in order to reproduce the measured average volumetric rate and velocity data. An average abdominal aortic blood flow rate of 3500 ml/min was introduced at the extended aortic inlet boundary as the inlet boundary condition. It was assumed that the kidney would maintain constant blood flow. Zero relative gradient pressure in the flow direction was used at both left and right renal artery outlets at all times. All of the above boundary conditions ensured to meet the primary conditions of linear equations. Transient Navier-Stokes governing equations were solved by ICEM CFX on the basis of the finite volume method.5、For qualitative assessment of MRA Images, image quality was based on the degree of whole-artery visualization, motion artifacts, and signal suppression of the background. The data was analyzed jointly by two radiologists with 11 and 5 years of experience in cardiovascular imaging. Both readers were blinded to the patient clinical data and a consensus was reached for each patient. The image quality of overall main renal artery MIP (Maximum intensity projection) images was graded by using a five-point scale:excellent. good, fair, poor, and non-diagnostic. Quantitative analysis was performed to determine diameter stenotic grades involving the main renal arteries. The extent of RAS was graded as follows:normal, mild, moderate, severe and occlusion.Maximum velocity of CFD results were compared with PSV from US which was used as a reference standard in 12 patients. As for CFD results in patients, the translesional PD was measured from the start of main renal artery to the end of stenosis, while the maximum velocity was measured at the most severe stenotic part. In normal volunteers, the PD was evaluated from the start to the end of main renal artery, while the velocity was calculated in the middle part of main renal artery. All MRF profiles were measured at the end of main renal arteries. 6、The maximum velocity measured using CFD was compared to PSV of US. Reliability was assessed using intraclass correlation coefficient; and Paired t test was used to measure statistically significant difference. For continuous variables, Paired t test and One-Way ANOVA (Fisher test) were applied to evaluate PD, velocity and MFR of various main renal artery stenoses and compare them with the control normal group.Results1、The quality of the images in all patients and volunteers were acceptable with quality scores of 4 (36 cases) and 3 (4 cases), and all CFD iteration process reached convergence to criteria of residual target 0.0001.2、The number of stenotic main renal arteries was 30 (left 15, right 15). Various degrees of RAS (minimum,2.0%; maximum,75.5%) were evaluated. Grade 1 was appreciated in twelve renal arteries, grade 2 in seven renal arteries, grade 3 in nine renal arteries, grade 4 in two renal arteries, and grade 5 in none of the renal arteries.3、Maximum velocity of CFD and PSV of US showed high intraclass correlation coefficient (value 0.955) in those 12 patients. And no significant difference (p>0.05) between them existed. The CFD results showed the same relationship between velocity and stenostic degree as the Doppler US did.4、As to normal volunteers, physiological and hemodynamic characters of main renal arteries were assessed. The average PD of left renal artery extended from 1.80 mmHg to 9.24 mmHg (mean 4.33±1.96 mmHg) and right between 1.82 mmHg and 8.45 mmHg (mean 4.11±1.94 mmHg). The maximum velocity of left normal main renal artery was between 0.67 m/s and 1.09 m/s (mean 0.82±0.12 m/s), and right between 0.64 m/s and 1.18 m/s (mean 0.80±0.16 m/s). MFR values ranged from 541.7 ml/min to 814.7 ml/min (mean 653.7±108.1 ml/min) on the left and from 385.3 ml/min to 658.3 ml/min (mean 546.3±108.1 ml/min) on the right. The velocity, PD and MFR of main renal artery were compared between right and left renal arteries, and these values showed no significant difference (p>0.05).5、The different grade of stenosis evoked different PD. The PD profiles were much higher in grade 3 and 4 stenotic cases than in normal, grade 1 and 2 cases. The mean PD, relevant to different stenotic grades of both left and right, were calculated as 4.22±1.90 mmHg for normal cases,5.86±3.41 mmHg for grade 1,14.60±6.30 mmHg for grade 2,68.36±31.01 mmHg for grade 3, and 120.50±2.71 mmHg for grade 4, respectively. No significance existed among normal, grade 1, and grade 2 cases (p>0.05). However, all of them were significantly different from grade 3 and grade 4 (p<0.001). Measurements in grade 4 were significantly higher than those in grade 3 (p<0.001).6、The maximum velocities increased with the severity of stenotic degree. And the corresponding velocities of different grades of RAS were:0.94±0.23,1.56±0.41, 3.43±0.98, and 4.65±0.31 m/s (maximum 4.87 m/s of diameter stenotic degree 75.5%) for grade 1,2,3 and 4. There was a negligible difference between the velocity in normal renal arteries and grade 1 stenotic vessels (p=0.059). Statistical significance in velocity existed between grade 2-4 and control group (p<0.001), and also among grade 2, grade 3, and grade 4 (p<0.01). A spike of more than 2.0 m/s for stenotic degree of 50% and about 4.0 m/s for 70% stenosis was observed.7、The general tendency of MFR decreased with the stenotic degree of RAS. Mean MFR of unilateral main renal artery were 544.4±132.8,502.1±122.6,299.0±122.6 and 137.4±5.0 ml/min, for grade 1,2,3 and 4 cases, respectively. A non-significant trend existed among normal, grade 1 and grade 2 (p>0.05). No significant difference was observed between grade 3 and grade 4 (p>0.05). However, the two groups showed different statistical significance (p<0.01), namely, MFR significantly changed in cases with more than 50% stenoses.ConclusionThis preliminary study shows that unenhanced-MRA-based CFD can be utilized to noninvasively analyse hemodynamic parameters of RAS. The acquired variables may provide meaningful information regarding of the relationship between RAS and hemodynamics. for grade 2,68.36±31.01 mmHg for grade 3, and 120.50±2.71 mmHg for grade 4, respectively. No significance existed among normal, grade 1, and grade 2 cases (p>0.05). However, all of them were significantly different from grade 3 and grade 4 (p<0.001). Measurements in grade 4 were significantly higher than those in grade 3 (p<0.001).6、The maximum velocities increased with the severity of stenotic degree. And the corresponding velocities of different grades of RAS were:0.94±0.23,1.56±0.41, 3.43±0.98, and 4.65±0.31 m/s (maximum 4.87 m/s of diameter stenotic degree 75.5%) for grade 1,2,3 and 4. There was a negligible difference between the velocity in normal renal arteries and grade 1 stenotic vessels (p=0.059). Statistical significance in velocity existed between grade 2-4 and control group (p<0.001), and also among grade 2, grade 3, and grade 4 (p<0.01). A spike of more than 2.0 m/s for stenotic degree of 50% and about 4.0 m/s for 70% stenosis was observed.7、The general tendency of MFR decreased with the stenotic degree of RAS. Mean MFR of unilateral main renal artery were 544.4±132.8,502.1±122.6,299.0±122.6 and 137.4±5.0 ml/min, for grade 1,2,3 and 4 cases, respectively. A non-significant trend existed among normal, grade 1 and grade 2 (p>0.05). No significant difference was observed between grade 3 and grade 4 (p>0.05). However, the two groups showed different statistical significance (p<0.01), namely, MFR significantly changed in cases with more than 50% stenoses.ConclusionThis preliminary study shows that unenhanced-MRA-based CFD can be utilized to noninvasively analyse hemodynamic parameters of RAS. The acquired variables may provide meaningful information regarding of the relationship between RAS and hemodynamics.Part ThreeAnalysis of Various Simulated Angioplasty of Renal Artery Stenosis in Hypertension Using Computational Fluid Dynamics TechnologyObjective Clinicians remain in sore need of better tools to identify when renal function can be or no longer be improved with renal revascularization. And there is still lack of standard methods to assess hemodynamic or functional reserve of renal artery revascularization. The purpose of this study was to assess predictive ability of MR angiography-based computational fluid dynamics for the hemodynamic indexes of renal artery stenosis of hypertensive patient after simulating angioplasty and stenting.Materials and Methods1、The tested patient with a history of hypertensive medication for 20 years had a current blood pressure 150/95mmHg. The right RAS was diagnosed with 70% diameter stenosis and with normal kidney size using unenhanced MRA.2、A commercial segmentation software Mimics was utilized to transfer MRA images (DICOM format) into vessel-shape geometric data (STL format). Following the patient-geometry segmentation, a number of different idealized models of typical stents were generated using Pro Engineer Wildfire, and the patient-geometry and the idealized stent geometries were merged together to form a new patient-geometry. The resulting three-dimensional STL formatted artery surface data was meshed using ICEM CFD to generate a suitable computational grid for CFD simulation. Navier-Stokes governing equations were solved using ANSYS CFX.3、Seven modified models were reconstructed with different diameter stenotic degrees from 60% to 0 with an interval of 10%. The hemodynamic parameters were analyzed for total eight stenotic degrees, including the original 70% stenosis.Results1、The translesional PD of right stenotic main renal artery of different modified simulating stenoses are shown along with the PD of left normal renal artery. The PD profile of modified stenosis was much higher than that of normal left side. The translesional PD of right main renal arteries undulately decreased with the declining severity of stenotic degrees. The PD were calculated as 79.78 mmHg for original 70% stenotic degree,59.16 mmHg for 60%,58.99 mmHg for 50%,51.99 mmHg for 40%, and 36.08 mmHg for 30%,33.62 mmHg for 20%,29.37 mmHg for 10% and 22.12 mmHg for modified non-stenotic model, respectively.2、According to the patient’s blood pressure 150/95mmHg measured during and after long-term medication, the systolic blood pressure of abdominal aorta decreased and the pressure percentage variation calculated as 13.7%,14.7%,18.0%,24.3%, 28.0%,29.2%,34.0% with the decreasing modified diameter stenotic degree from 60% to 0.3、The results indicated that the mean velocity inside renal arteries undulately decreased with the decline of modified stenotic degrees. The maximum velocity of 70% degree stenosis was 4.65m/s. The mean velocities of different degrees were:3.68, 2.84,3.10,2.73,2.09,1.80,1.95,1.64 m/s for stenotic degree 70%,60%,50%,40%, 30%,20%,10% and 0, respectively.4、The general tendency of MFR of ipsilateral renal artery increased with the stenotic degrees. MFR of right renal artery were calculated 178.1,228.2,338.4,472.8, 693.4,528.3,710.5,722.8 ml/min, for successive decreasing stenotic degrees.ConclusionOur initial application of CFD technology on modified renal artery stenosis demonstrates its potential for an improved prediction of the hemodynamics of renal artery after simulated percutaneous renal angioplasty and stenting. CFD based on unenhanced MRA may non-invasively help to guide the clinical practice of renal artery stenosis intervention.