RPS 1415 - Advances in vascular imaging

Purpose:

Perfusion decellularisation is a promising method for the generation of non-immunogenic organs from allogeneic and xenogeneic donors. A number of imaging modalities are used to assess vascular integrity in bioengineered organs, including fluoroscopic angiography, with no consistency in the method used. None of these techniques addressed changes in permeability, while the demonstration of patency was limited to visualisation of arterial tree. The goal of this study was to evaluate the use of fluoroscopic angiography performed under controlled flow conditions for the assessment of vascular integrity in bioengineered kidneys.

Methods and materials:

Porcine kidneys underwent ex vivo fluoroscopy before and after perfusion decellularisation, under controlled flow conditions. Arterial and venous patency were defined as the visualisation of contrast medium (CM) in distal capillaries and the renal vein, respectively. For permeability assessment, grey-scale intensity within the parenchyma was measured and the washout index (W_index) was calculated at several time points during a washout phase. Extravasation of CM was documented following iatrogenic injury.

Results:

No differences in patency were detected in decellularised kidneys compared with native kidneys. However, a significantly lower W_index was calculated for decellularised kidneys, indicating a delayed CM clearance and increased vascular permeability. Focal opacities representing extravasation of CM into the parenchyma were only detected in decellularised kidneys. Iatrogenic leakage was equally detectable in both groups, with DSA scans revealing even minimal CM leakage.

Conclusion:

Quantitively assessment of permeability should be coupled with patency when studying the effect of perfusion decellularisation on kidney vasculature. Flow-controlled fluoroscopic angiography based on the proposed methodology is an accessible, accurate, and sensitive method that should be adopted as the method-of-choice for vascular assessment in bioengineered organs.

Limitations:

Larger sample sizes are needed to confirm our findings.

Ethics committee approval

Performed in accordance with IRB guidelines.

Funding:

No funding was received for this work.

Purpose:

Altered blood flow and wall shear stress (WSS) in arteriovenous fistulae (AVF) might make the access stenosis, causing the dialysis procedure to become ineffective. For our study, the stenosis of AVF is considered in three specific sites: the anastomosis (Type-I), the vein straightens out after the curved region (Type-II), and the curved region of the swing segment (Type-III). WSS in these sites could be measured by V Flow, which is a high-frame-rate ultrasound vector flow imaging (VFI) implemented based on multi-directional Doppler techniques.

Methods and materials:

This study established the first quick measurement of vascular WSS by V Flow in AVF (available on a clinical system, Resona 7 manufactured by Mindray, Shenzhen, China). 15 haemodialysis patients were included; 9 males and 6 females. The dialysis age was 38.0±25.1 months. The mean age of the fistula was 21.4±18.6 months. Each fistula was functioning for dialysis access at the time of our examination.

Results:

The level of mean WSS in Type-I, -II, and -III sites were 0.9±0.6Pa, 1.2±0.5Pa, and 0.8±0.4Pa. The level of max WSS was 2.7±1.3Pa, 4.2±2.5Pa, and 1.7±0.7Pa in these sites. It was found that the WSS level of Type-III was significantly lower than the others (P<0.05), except the mean WSS in Type-I and -III (P=0.6). The level of fistulae flow rates was about 559±126.5 ml/min measured by V Flow.

Conclusion:

The ultrasound VFI technique is convenient to study the changes of WSS and flow in AVF. The lowest WSS in Type-III might explain that this site is the most susceptible site for stenosis. WSS is expected to be a sensitive and non-invasive method for the early prediction stenosis of AVF.

Limitations:

A pilot study.

Ethics committee approval

Approved by the Hospital Review Board.

Funding:

Supported by Hospital funding YN2017QN04.

Purpose:

Doppler ultrasound (DUS) is routinely performed for vascular mapping prior to haemodialysis arteriovenous fistula (AVF) creation but has the disadvantage of not visualising the central vasculature. Ferumoxytol, an iron oxide nanoparticle, provides an alternative to gadolinium contrast for magnetic resonance angiography (MRA). We assessed the clinical utility of ferumoxytol-enhanced MRA (FeMRA) before autogenous upper limb AVF creation compared with DUS.

Methods and materials:

In a prospective comparative study, paired FeMRA and DUS were obtained the same day. Vessels were evaluated for diameter, stenosis or occlusion, arterial disease, and central stenosis by independent readers. Interclass correlation coefficients (ICC) and Bland-Altman plots examined inter- and intra-reader variability. Based on accepted standards for AVF creation, an algorithm was designed to predict the AVF outcome relying on mapping findings. Logistic regression models were created with AVF success as the dependent variable and age, sex, and DUS or FeMRA mapping as the independent variables.

Results:

59 patients (mean age 59 years) had FeMRA and DUS, and 51 AVF were created. FeMRA showed excellent inter- and intra-reader repeatability (ICC 0.84-0.99). In addition to identifying 15 central vasculature stenoses, FeMRA explicitly characterised anatomical abnormalities and variants in arm vessels. On multivariable regression analyses, FeMRA mapping independently predicted AVF success in models including [OR: 6.49 (95% CI 1.70-24.79); p=0.006] and excluding central stenoses [OR: 4.58 (95% CI 1.25-16.83); p=0.02].

Conclusion:

FeMRA better predicted the AVF successful outcome compared to DUS and has the added advantage of identifying central vessels pathology.

Limitations:

The translation of study protocol to clinical practice. Ferumoxytol is not yet licensed as a contrast agent for MRI and used off-label.

Ethics committee approval

The study was approved by the institutional review board [Research Ethics Committee (REC) reference: 16/NS/0099, NCT02997046].

Funding:

No funding was received for this work.

Purpose:

To introduce a new approach for aneurysm coiling by using a magnetic micro-robot, which is steered by the magnetic fields of a magnetic particle imaging (MPI) scanner in an in vitro study.

Methods and materials:

For in vitro phantom experiments, a high precision model of a middle cerebral artery was used, which was 3D-printed and filled with water. For the micro-robot, we used a 3D-printed helically shaped swimmer, which can be navigated by magnetic fields. The required homogeneous magnetic fields of low amplitude (about 10 mT) and rotating field vector are generated by a commercial MPI scanner (Bruker BioSpin, Ettlingen, Germany). Low rotation frequencies (around 10 Hz) are applied and the swimmer follows the rotation. Due to the swimmers’ shape, a forward movement is induced. The magnetic fields are pre-calculated according to the pathway the swimmer has to cope. The swimmer has a length of 3.0 mm and a width of 1.2 mm, fitting the blood vessels of the phantom. It is coated with magnetic nano and microparticles to introduce magnetic properties. The magnetic nanoparticles could serve as tracer material for visualisation with MPI.

Results:

The swimmer can be navigated by external magnetic fields through a blood vessel towards an aneurysm within a vessel phantom. The magnetic fields were applied with an MPI scanner, which allowed the visualisation of the swimmer.

Conclusion:

The first in vitro experiments demonstrate the feasibility of steering a micro-robot by pre-calculated magnetic fields of an MPI scanner through a vessel phantom towards an aneurysm, which paves the way for an untethered coiling procedure.

Limitations:

In vitro experiments were performed without flow.

Ethics committee approval

n/a

Funding:

The Federal Ministry of Education and Research, Germany (BMBF), Grant No. 13GW0230B (FMT).

Purpose:

To evaluate the heating of a recently developed re-dilatable stent for the treatment of aortic coarctation in neonates and small children during magnetic particle imaging (MPI) and magnetic resonance imaging (MRI).

Methods and materials:

All measurements were performed using a recently developed cobalt-chromium stent (BabyStent, Osypka) with a new stent design which allows for re-dilatation and adjustment of the diameter from 6 to 16 mm. At a diameter of more than 12 mm, the stent loses radial integrity while fracturing at dedicated breaking points. Temperature measurements were performed with fibre-optic thermometers during 7-minute MPI and MRI scans. 6 different stent diameters (6, 8, 10, 12, 14, and 16 mm) were evaluated in this study.

Results:

In MPI, all stents with continuous struts showed a temperature increase (max. 2.3 K). The measured temperature differences increased with growing diameters up to 12 mm, whereas the stents with discontinuous struts at 14 and 16 mm showed nearly no heating. In contrast to MPI, the investigated stents showed no heating during the MRI measurements.

Conclusion:

The stents showed no heating during MRI and only clinically insignificant heating during MPI and can safely be examined with both modalities. However, in MPI, the stent diameter and the strut design influence the temperature difference.

Limitations:

The absence of flow could be a potential limitation of our study. To prevent potential cooling effects, we consciously created a worst case scenario study with static conditions and stents surrounded by air.

Ethics committee approval

n/a

Funding:

The German Federal Ministry of Education and Research BMBF, grant number 13GW0071D.

Purpose:

Patient-specific inlet artery flow has been applied in intracranial aneurysm simulations. This study aims to identify the intra-individual flow variation in phase-contrast MR (PCMR) and how it influences the following computational haemodynamics.

Methods and materials:

2D PCMRs were performed in a longitudinal patient cohort of intracranial aneurysms to determine the inlet flow boundary condition. The intraclass correlation coefficient (ICC) and the coefficient of variance (CV) across 3 time points were used to identify reproducibility and variation. Computational fluid dynamics were performed using the original flow and flow with the corresponding variation, respectively. The wall shear stress (WSS), low shear area (LSA), time-averaged wall shear stress (TAWSS), and oscillatory shear index (OSI) were compared.

Results:

51 patients (14 males; mean age: 62 years; mean follow-up interval: 7.2 months) were included. The reproducibility of the flow and velocity was good, with ICCs of 0.71-0.90. A 10% CV of mean flow was identified. The variation of flow data was smaller than the velocity data (p<0.05). The variation of flow and velocity at the end-diastolic phase was larger than that of the mean and systolic phase (p<0.01). A 10% of flow variance would lead to haemodynamic changes including 20% of WSS at the systolic phase, 19.6% of WSS at diastolic phase, 5.0% of LSA at systolic phase, 5.2% of LSA at diastolic phase, 20.1% of TAWSS, and 4.6% of OSI.

Conclusion:

The flow should be preferred rather than the velocity in choosing inlet boundary conditions due to a smaller variance. A 10% of flow variation could lead to a series of haemodynamic changes. LSA is a more stable parameter than WSS.

Limitations:

n/a

Ethics committee approval

Under IRB approval of UCSF. All subjects gave informed written consent.

Funding:

No funding was received for this work.

Purpose:

To demonstrate the feasibility and reliability of flow velocity measurements using time-resolved CT-angiography in a dynamic flow phantom.

Methods and materials:

All experiments were performed with a single lumen flow phantom with a 6 mm inner diameter. Saline was infused at varying injection rates with a contrast injector (Dual Shot GX 7-Nemoto) and the contrast bolus (2 mL iomeprol 350 at 0.33 mL/s) was subsequently injected using a syringe injector (Alaris GH Syringe Pump). The two injectors were connected to a silicone tube which was contained within a solid 2% agar-agar solution to mimic soft tissue.

Saline was injected at 6 flow rates to achieve the following reference velocities: 21.2 mm/s, 38.9 mm/s, 60.1 mm/s, 81.4 mm/s, 99.0 mm/s, and 120.3 mm/s. For each velocity, 6 consecutive time-resolved CTA acquisitions were performed (Revolution CT, GE Healthcare), each consisting of 20 repeated axial acquisitions with an interscan delay of 1.2s at 160 mm (256x0.625 mm) collimation. The mean centerline flow velocities were automatically assessed and differences and correlations with reference settings were evaluated with the Wilcoxon signed-rank test and the Pearson correlation.

Results:

From the CT image data, the average obtained velocities were 21.6±5.6 mm/s, 34.5±6.8 mm/s, 61.5±3.1 mm/s, 76.3±14.5 mm/s, 104.9±27.0 mm/s, and 143.6±42.1 mm/s. No significant difference was observed with the reference velocities (p=0.912) and a significant linear correlation was found (r=0.903, p<0.01). Up to 99.0 mm/s, the measurement error was confined within 12%.

Conclusion:

The results of this study suggest that mean flow velocities within the range of typical blood flow velocities (40 mm/s-70 mm/s) can be accurately measured with high precision in a 6 mm flow phantom using time-resolved CTA.

Limitations:

Of minor importance.

Ethics committee approval

n/a

Funding:

Fonds Wetenschappelijke Onderzoek Vlaanderen (FWO).