The Effect of Position on Arterial and Venous Geometry in the Arm: a Single Case Observational Study

Stephen P Broderick1, Stephen J Gandy3,4, Leonard D Brown1, Efstratios Kokkalis2, J Graeme Houston2,3, Michael T Walsh1

1 Centre for Applied Biomedical Engineering Research (CABER), Department of Mechanical, Aeronautical and Biomedical Engineering (MABE), Material and Surface Science Institute (MSSI), University of Limerick, Limerick, Ireland2 Cardiovascular and Diabetes Medicine, Ninewells Hospital and Medical School, University of Dundee, Dundee, UK3 Department of Clinical Radiology, Ninewells Hospital, National Health Service (NHS) Tayside, Dundee, UK.4 Department of Medical Physics, Ninewells Hospital, National Health Service (NHS) Tayside, Dundee, UK.

The suitability of a subjects arm for arteriovenous fistulae creation is often examined byultrasonic imagining. However, while this method is sufficient for a diagnostic methods, theabsence of 3-Dimensional geometry does not facilitate further investigation by means ofComputational Fluid Dynamics (CFD), which illuminates more information, such as wall shearstress, velocity profiles and flow structures. Magnetic Resonance Imaging (MRI) provides such3D information, but is more spatially restrictive than Ultrasound. Given the spatial restrictionof the MRI bore, few positions are available for scanning the arm. As no ideal position isdetermined, the effect on vascular geometric structure is examined in a single case andcommented on.

Introduction

Figure 1: Illustration of the approximate positions of arm and person within the MRI bore, with scanning area

indicated by the grey bounding box. “The supine” position (a) and a prone position labelled “the superman” (b)

were the two chosen configurations.

A healthy subject was scanned in a Siemans MAGNETOM Trio MRI scanner in two positions.Firstly, lying prone with the arm outstretched above the head and secondly lying supine with the armdown by the side. This is illustrated as “the superman” (b) and “the supine” (a) position in figure 1.The scanning area was of the upper arm focusing on the brachial artery, basilic and cephalic vein.Two MRI scanning sequences were used. These were the 2D time-of-flight (2D TOF) and the true-FISP with parameters indicated in table 1.

Scan data were imported into 3D Slicer 4.3.1. Vessel geometry was segmented manually in each slicebased on image contrast and slice continuity. The humerus bone was also reconstructed and used as areference to align both scan positions using an Iterative Closest Point (ICP) algorithm in MeshLab1.3.3. Cross-sections of the brachial artery and cephalic vein were taken at approximately 5 mmintervals along vessel path, with origin slices at physiological landmarks such as vessel junctions.

Vessel comparisons were performed for statistical significance using a paired t-test between bothpositions using slices at the same relative position. Diameter was estimated from area by assuming acircular cross-section.

Methods

Scan type True FISP 2D Time of Flight multi-slab

Slice thickness 1 mm 1.5 mmRepetition Time 6.26 13Echo Time 3.13 5.8No. of Averages 1 2Magnetic Field Strength 3 T 3 TPixel Bandwidth 558 164Flip Angle 70° 18°Field of View 120*120 140*140Matrix Size 256*256 512*512No. of Slices 160 100

Table 1: Parameter list for the MRI sequences

Figure 2: Slices from the upper arm to the elbow from left to right of the two MRI sequences. Of interest are the

various vessels, with the brachial artery shown beside the red * symbol, the cephalic vein with the blue square and

basilic vein with the blue triangle. The remaining vessels are vein tributaries of the aforementioned veins.

True FISP

2D Time-of-Flight

** *

** *

Figure 3: Shown on the left is the medial aspect of the upper arm with the exploded view illustrating the three major vessels of interest, the cephalic vein (a), basilic vein

(b) and the brachial artery (c). These vessels are shown in more detail on the right side for both the supine and superman positions. In (a), both positions mirror similar

shape, size and tertiary vessels, with the superman appearing less stretched but with marginal narrowing in the top left branch. This is in contrast to (b) where much of the

detail is absent in the supine position. Vessels appear as joined to neighbours and the primary vessel is significantly narrowed. In (c) the diameter appears similar but the

superman displays a helical profile as evident from the front and side view combined, while in the supine position, this is not present.

Supine Superman

(a)

(b)

(c)

(a)

(b)

(c)

0

2

4

6

8

0

2

4

6

8

0

2

4

6

8

Ves

sel D

iam

eter

(m

m)

(i)

(ii)

(iii)

Figure 4: Plots of vessel diameter along the vessel length for both poses. Brachial artery (i)

cephalic vein (ii) and the cephalic tributary to vein confluence are shown in solid and dotted

lines indicating the superman and supine positions respectively. To the right of the graphs

are the slice numbers and calculated p-value from a 95% contrivance interval paired t-test

Superman

Supine

N = 26

p < 0.057

N = 25

p < 0.049

N = 13

p < 0.005

Discussion and ConclusionVessel shape and diameter has been shown to change with arm position. MRI scanning was performed using twosequences, the 2D TOF and True FISP as shown in figure 2. The True FISP was chosen as a more reliable sequence asthe 2D TOF showed partial corona shadowing on vessel and was more prone to noise in smaller vessels. The volunteerwas scanned in the two positions as shown in figure 1. Examining the vessels in figure 3(a), the supine is more stretchedthan the superman, but centreline profiles appear similar. The distal section of the superman (left,top) is somewhatcompressed, which might be caused by head or pillow pressure. In (b) the basilic vein is irregular and absent in sectionsfor the supine position which contrasts with the full form of the superman position. As the coil is wrapped around thearm, combined with the minimal space for the body and arm within the MRI bore suggests pressure was applied on thevein collapsing it. The superman position appears full as pressure was not applied to it and the vein may have filledgreater due to restriction of return flow at the shoulder due to the position. The artery (c) showed little change indiameter but displayed a helical form in the superman. Figure 4 examines the change in diameter over the vessel length.The artery (i) shows little change and was not considered significant. This acts as a control, ensuring that changes are notsome inherent MRI parameter. Only the cephalic vein was plotted as the basilic was incomplete for the supine position.Some changes were noted in (ii) and was modestly significant while (iii) was deemed a significant difference between themeasurement of both positions. Differences appeared most notable distal of the vein confluence.With clinical measures and base line wall shear CFD calculations based off of MRI reconstructions, awareness must begiven when examining collapsible structures in pre and post arteriovenous junction creation.

This project has received funding from the EuropeanUnion's Seventh Framework Programme for research,technological development and demonstration undergrant agreement no 324487