A complex flow phantom for medical imaging : ring vortex phantom design and technical specification

TL;DR: A novel, cost-effective, portable, complex flow phantom is proposed and the design specifications are provided, which employs a piston/cylinder system for vortex ring generation, coupled to an imaging tank full of fluid, for vortex propagation.

Abstract: Cardiovascular fluid dynamics exhibit complex flow patterns, such as recirculation and vortices. Quantitative analysis of these complexities supports diagnosis, leading to early prediction ...

Summary

1. Introduction

• The fluid dynamics of the cardiovascular system are associated with many complexities and this has provoked interest from scientists for decades.
• Magnetic Resonance angiography and Doppler Ultrasound are examples of technologies that offer valuable flow-based assessment of the cardiovascular system.
• Blood speckle imaging is available on the GE Healthcare Vivid E95TM (GE Healthcare, Illinois, USA).
• Manufacturers are not clear about tolerances and accuracy specifications [18] and the lack of well-defined protocols makes the Quality Control of scanners that measure flow and velocity challenging [15,16].
• With its uniquely controllable characteristics, the ring vortex is a suitable flow reference for a flow test object.

2. Materials and Methods

• The complex flow phantom design described here uses a piston to propel a slug of fluid along a cylindrical channel, through an orifice port that connects to an open tank of fluid (containing water, Blood-Mimicking Fluid etc.) where propagation of the ring may be observed and imaged.
• A vortex ring naturally forms when a column of fluid is pushed through an orifice into a neighbouring expansive fluid environment.
• The aﾉ┌ｷS ヮヴﾗヮWﾉﾉWS デｴヴﾗ┌ｪｴ デｴW ﾗヴｷaｷIW さヴﾗﾉﾉゲ ┌ヮざ デｴW ﾗヴｷaｷIW face, forming a toroidal core of vorticity that may ultimately detach (at high enough Reynolds number) and propagates along the axisymmetric axis of the ring.
• The whole process is managed by computer control of piston displacement that delivers the fluid slug with a precisely specified velocity/displacement profile.

2.1 Imaging Tank

• The imaging tank is the principal component of the phantom since it is the environment in which the ring vortex propagates; it is the volume used to image the generated flow.
• The tank is manufactured from clear poly(methyl methacrylate) (PMMA) which rests on four small screw feet.
• These are adjustable to allow levelling of the system via two spirit levels placed on opposite walls of the tank box.
• Several reference markers and a ruler are laser-cut into the surfaces of the tank walls to support positioning and measurement.
• The top of the tank is open, offering easy access for Ultrasound imaging probes.

2.2 Piston Cylinder System

• The generation of vortex rings relies on three components: a cylindrical channel, a plunger and an orifice that connects the channel to the imaging tank.
• The cylindrical channel used here is of internal diameter 70+/-0.10 mm, reamed from a solid cylinder manufactured from PMMA .
• One end of the chamber includes a screw thread for leak-proof coupling with the orifice and the imaging tank.
• The piston head of the plunger incorporates an O-ring to provide a leak-proof seal between piston and the cylindrical chamber .
• These generate rings of several centimetres diameter with characteristics that are relevant to human physiology.

2.3 Programmable Actuator

• The nut/lead screw arrangement of the piston enables it to be driven by a linear stepper motor - this constitutes the actuator of the ring vortex generator.
• A Nema 23 external linear stepper motor drives the threaded screw of length 150 mm (Nema 23 external linear actuator, OMC Corporation Limited, Nanjing, China).
• Power is delivered through a digital stepper driver (OMC Corporation Limited, Nanjing, China) which has connections to a switching 150 W power supply (OMC Corporation Limited, Nanjing, China).
• The digital system relies on CMOS logic (0-5V), with a square-wave signal of varying frequency (from 0.5 kHz to 1 kHz) but fixed duty cycle (50%) controlling the digital motor displacement.
• The total cost of all the electronic components is less than one hundred Euros at the time of writing (2019).

2.4 The Assembled System

• The interconnection of these components provides the system depicted in Figure 5.
• The actuator is programmed to deliver a preconfigured displacement profile (eg. top-hat) that moves the piston through the cylinder and propels a slug of fluid through the orifice (e.g. 3 cm3 of displaced fluid in 50 ms).
• This results in a controlled, propagating ring vortex of an orificedependent size that travels steadily (with velocities < 1 ms-1) along the length of the tank .
• Images captured by an imaging system can be referenced to the features of the flow, permitting characterisation of imaging system performance.
• Bulk features (ring speed, size) are readily established through simple optical/video methods [20] whereas finer details of the flow benefit from technologies such as particle image velocimetry (PIV) or even computational fluid dynamics (CFD) [21].

3. Demonstration and Application

• Reynolds number calculations are based on flow through the orifice generating the ring (Re = Vd/; - density, V-velocity, d-diameter, -viscosity).
• In order to confirm the functionality of the phantom design, two different measurement methods were used to characterise the generated flows, namely optical/video measurements and Laser-PIV; details are provided in sections 3.1 and 3.2 below.
• Optical/video measurements were accomplished on the premises of Leeds Test Objects Ltd (Leeds Test Objects Ltd, Boroughbridge, United Kingdom) as previously described by Ferrari et al [20].
• The PIV methodology followed that described by Wieneke [24], with calibration procedures performed under the supervision of a LaVisionUK Ltd (LaVisionUK Ltd, Bicester, United Kingdom) application consultant.

3.1 Optical/video acquisition

• Vortex progress was captured at 25 frames per second using a Sony camera HDR-PJ220E (Sony Corporation, Tokyo, Japan).
• The camera was placed on a tripod at a distance of 3.8 m from the phantom, obtaining a clear view through the sides of the transparent tank.
• This zone was chosen as the region of interest for all measurements (including PIV).
• In addition, a 1/100sec universal counter-timer (RS Components Ltd, Stock No 612-445) was placed within the field of view of the images to provide visible timing data in support of the measurements.
• Distance measurements from still frames enabled characterisation of vortex ring position and translational speed.

3.2 Laser-PIV acquisition

• The Laser-PIV setup required neutrally buoyant scattering particles mixed throughout the tank/cylinder system to provide satisfactory conditions for flow capture using the stereo Laser-PIV technique.
• Fluorescent particles of size 10-20 micron were mixed within the volume of water, which was illuminated by a laser sheet that cut the propagating ring through its centre in the vertical plane.
• Reconstruction of particle displacement using two camera projections (stereoscopic view) and digital image correlation allowed the reconstruction of the flow field [24] at a spatial resolution of 0.4 mm and temporal steps of 0.071 s (14 Hz).
• Accuracy of the instantaneous velocities of the PIV system is declared by LaVisionUK Ltd (LaVisionUK Ltd, Bicester, United Kingdom) to be better than +/-0.1%.
• Mapping the flow field of the Vz components at each frame and knowing the frame rate of the acquisition (14 fps), the translational velocity of the ring was calculated.

4.1 Optical/Video measurements

• Average vortex ring translational velocities (Vtrans with standard deviation expressed by the error bars) as a function of vortex ring position are plotted in figs 6,7 (see Table 1 for the configurations tested).
• For improved clarity, results for the configurations that produced faster ring vortices (from 15 cm/s to 80 cm/s) are shown in Figure 6 while results from slower configurations (from 0 cm/s to 15 cm/s) are shown in Figure 7.
• In order to compare the optical data directly with the Laser-PIV translational velocity measurements, each selected configuration (Table 1) corresponds to a specific colour and marker-shape combination.
• In each case, variability within 10 runs is presented as +/-1SD on the plots and cited as coefficient of variation (%) within the text.
• Such errors calculated with the optical/video method are always less than +/-10%.

4.2 Laser-PIV

• For consistency, average PIV vortex ring translational velocities with standard deviation are also plotted (dashed lines) as a function of vortex ring position .
• These are calculated from 10 vortex ring acquisitions for each of the settings listed in Table 1 (as for the optical/video method).
• Relative errors were calculated as for the optical/video experiment, and the varying configurations confirm very similar behaviour to the optical results.
• Notably, the error increases for slower vortex ring configurations (e.g. Configuration 10, Table 1).

5. Discussion

• Bulk flow performance has been assessed using two identical systems, undertaken at two different places (Leeds and Sheffield), two weeks apart.
• In respect of ring vortex performance, Gharib observed that vortex ring generation depends on a non-dimensional parameter, described [25] as formation time or formation number.
• Importantly, the measurement processes described in this paper indicate that realistic tolerances for these reproducible flows can be asserted に hence the utility of this design as a complex flow phantom.
• With a longer-term ambition to compare flow imaging performance between different flow medical imaging modalities (Ultrasound, MRI and CT) - both for clinical and research purposes - the ring vortex phantom has been entirely manufactured from PMMA (except for the motor).
• In an environment where the quality control of flow medical imaging scanners is still challenging and confusing, this technology offers support for both research and clinical imaging activities.

6. Conclusion

• A technical specification for the design of a novel, cost-effective, portable complex flow test object currently compatible with ultrasound modalities has been provided.
• Early experience indicates that the technology has the characteristics suited to basic and more advanced ultrasound quality control checks at research and clinical level.
• The phantom as described, demonstrates that it can be reproducibly manufactured to deliver consistent performance within specified tolerances.
• The authors report no conflict of interest.
• This work is funded by the European Commission through the H2020 Marie Sklodowska-Curie European VPH-CaSE Training Network (www.vph-case.eu), GA No. 642612.

Figures

Figure 4a. Block-diagram of the ring vortex flow phantom, illustrating vortex propagation at velocity Vtrans

Figure 8: This plot present data for configurations 1, 4, 7, 9, 10 (table 1). The dispersion in ring speed is apparent from this representative collection of experimental runs illustrating ring position vs time. Ring velocity can be determined from line gradient, and variability <10% for each group (except for the 25mm orifice example). Notably, the rings of lower energy and speed are more variable. With a 2.5cm diameter orifice, and lower pump speeds, the ring might not even form.

Figure 4a. Block-diagram of the ring vortex flow phantom, illustrating vortex propagation at velocity Vtrans

Figure 5. Assembled ring vortex flow phantom.

Figure 7. Laser-PIV and coloured dye translational velocity measurements relevant to configuration 6 to 10 (Table 1). Each configuration corresponds to a particular marker/shape combination enabling direct comparison of optical/PIV measurements. Laser-PIV data are plotted with dashed lines while coloured dye data uses solid markers with bold lines. Error bars represent the standard deviation values on 10 acquisitions.

Table 1. Summary of the flow generating conditions, with Reynolds number at the throat of the orifice cited. Each configuration was repeated ten times to evaluate reproducibility

Figure 1. CAD drawing of the ring vortex tank.

Figure 9: Image (a) displays a colormap derived from the PIV data, depicting the y-component (vertical direction) of the fluid velocities within the ring vortex. (b) shows the structure of the ring (vertical crosssection) captured by the B-scan mode of a Doppler scanner. The color Doppler image (c) clarifies flow toward/away from the probe (placed at the top of the image) and apart from the persistence artefact, appropriately resembles the PIV data of image (a).

Figure 6. Laser-PIV and coloured dye translational velocity measurements relevant to configuration 1 to 5 (Table 1). Each configuration corresponds to a particular marker/shape combination enabling direct comparison of optical/PIV measurements. Laser-PIV data are plotted with dashed lines while coloured dye data uses solid markers with bold lines. Error bars represent the standard deviation values on 10 acquisitions.

Figure 3. CAD drawing of the attachable/detachable orifices.

Figure 2. CAD drawing of the piston cylinder system.

Full text

This is a repository copy of A complex flow phantom for medical imaging : ring vortex
phantom design and technical specification.
White Rose Research Online URL for this paper:
http://eprints.whiterose.ac.uk/149487/
Version: Accepted Version
Article:
Ambrogio, S., Walker, A., Narracott, A. orcid.org/0000-0002-3068-6192 et al. (3 more
authors) (2019) A complex flow phantom for medical imaging : ring vortex phantom design
and technical specification. Journal of Medical Engineering & Technology, 43 (3). pp.
190-201. ISSN 0309-1902
https://doi.org/10.1080/03091902.2019.1640309
This is an Accepted Manuscript of an article published by Taylor & Francis in Journal of
Medical Engineering and Technology on 15th July 2019, available online:
http://www.tandfonline.com/10.1080/03091902.2019.1640309
eprints@whiterose.ac.uk
https://eprints.whiterose.ac.uk/
Reuse
Items deposited in White Rose Research Online are protected by copyright, with all rights reserved unless
indicated otherwise. They may be downloaded and/or printed for private study, or other acts as permitted by
national copyright laws. The publisher or other rights holders may allow further reproduction and re-use of
the full text version. This is indicated by the licence information on the White Rose Research Online record
for the item.
Takedown
If you consider content in White Rose Research Online to be in breach of UK law, please notify us by
emailing eprints@whiterose.ac.uk including the URL of the record and the reason for the withdrawal request.

A Complex Flow Phantom for Medical Imaging: Ring Vortex Phantom
Design and Technical Specification
1,2,3
Ambrogio Simone,
3
Walker Adrian,
1,2
Narracott Andrew,
1,2
Ferrari Simone,
4
Verma Prashant,
1,2
Fenner John
1
Medical Physics, Mathematical Modelling in Medicine Group, Department of Infection, Immunity and Cardiovascular Disease, University
of Sheffield, Sheffield, UK
2
Insigneo Institute for In Silico Medicine, University of Sheffield, Sheffield, UK
3
Leeds Test Objects Ltd., Boroughbridge, UK
4
Medical Imaging & Medical Physics, Sheffield Teaching Hospitals NHS Foundation Trust, Sheffield, UK
Corresponding Author:
Simone Ambrogio
E-mail:
simoambr@gmail.com
(Orcid 0000-0003-1571-0016)
ABSTRACT
Cardiovascular fluid dynamics exhibit complex flow patterns, such as recirculation and vortices. Quantitative
analysis of these complexities supports diagnosis, leading to early prediction of pathologies. Quality Assurance
of technologies that image such flows is challenging but essential, and to this end, a novel, cost-effective,
portable, complex flow phantom is proposed and the design specifications are provided. The vortex ring is the
flow of choice because it offers patterns comparable to physiological flows and is stable, predictable,
reproducible and controllable. This design employs a piston/cylinder system for vortex ring generation, coupled
to an imaging tank full of fluid, for vortex propagation. The phantom is motor-driven and by varying piston speed,
piston displacement and orifice size, vortex rings with different characteristics can be produced. Two
measurement methods, namely Laser-PIV and an optical/video technique, were used to test the phantom under
a combination of configurations. Vortex rings with a range of travelling velocities (approximately 1-80cm/s) and
different output-orifice diameters (10-25mm) were produced with reproducibility typically better than +/-10%.
Although ultrasound compatibility has been demonstrated, longer-term ambitions include adapting the design
to support comparative studies with different modalities, such as MRA and X-ray-CTA.
Keywords: Medical Imaging Phantom, Doppler Ultrasound Quality Assurance, Vortex Ring, Complex Flow
Phantom, Laser PIV

1. Introduction
The fluid dynamics of the cardiovascular system are
associated with many complexities and this has
provoked interest from scientists for decades.
Recirculation, turbulence, jets and vortices are
observed both in healthy and pathophysiological
conditions [1,2]. Quantitative analysis of the
distribution of blood velocity patterns can support
diagnosis of the cardiovascular system, leading to
early prediction of pathologies, improvements of
surgical outcomes and evaluation of potential
therapies [2,3]. Magnetic Resonance angiography and
Doppler Ultrasound are examples of technologies that
offer valuable flow-based assessment of the
cardiovascular system. 4D Flow MRI is a powerful
technique that provides a time-resolved 3D velocity
field. Velocity is encoded along all three spatial
dimensions in the vessel of interest (4D = 3D + time).
Unfortunately, this technique is not clinically routine
because the acquisition time exceeds that demanded
by the clinical workflow [4]. Doppler Ultrasound is
portable, non-invasive, cost effective, does not involve
ionizing radiation and provides qualitative and
quantitative real-time information about volumes (i.e
stroke volume, ejection fraction), inferred pressures
(across heart valves) and flows (i.e. cardiac output)
[5,6]. For these reasons, Ultrasound is currently the
first choice as a diagnostic modality for the
assessment of several cardiovascular pathologies [7].
Recent developments include 2D and 3D real-time
angle independent Ultrasound Doppler imaging
techniques, based on Vector Flow Imaging (VFI), and
post-processing algorithms for particle velocimetry
tracking and volume quantification
[6,7,8,9,10,11,12,13]. A specific Ultrasound VFI
technique - Transverse Oscillation (TO) - has been FDA
approved (2013) and is clinically available on
commercial scanners (eg. BK Ultrasound, Nova Scotia,
Canada, and Carestream Health, Ontario, Canada) for
real-time analysis of complex flows in valves,
bifurcations and heart chambers [6,14]. A colour
Doppler based VFI technique is also commercially
available on clinical scanners manufactured by Hitachi
(Hitachi Ltd., Tokyo, Japan), GE Healthcare (GE
Healthcare, Illinois, USA) and Mindray (Mindray
Medical International Limited, Shenzen, China). Blood
speckle imaging is available on the GE Healthcare Vivid
E95
TM
(GE Healthcare, Illinois, USA). In this domain,
improved technologies for calibration of such medical
imaging techniques are essential. Complex flows are
needed as a standard reference for the validation of
innovative flow estimation algorithms.
Recent audits concerning Ultrasound Quality Control
protocols indicate that the current flow/velocity test
objects available on the market are rather limited,
expensive, and often fail to reproduce physiological
conditions and physiological waveforms
[15,16,17,18]. Manufacturers are not clear about
tolerances and accuracy specifications [18] and the
lack of well-defined protocols makes the Quality
Control of scanners that measure flow and velocity
challenging [15,16]. For example, experimental mean
velocity measurements using a commercial string
phantom (as recommended in IPEM Report 102,
Institute of Physics and Engineering in Medicine, 2010)
for linear, curvilinear and phased array probes
reported errors exceeding 20%, 50% and 40%,
respectively [17]. These values are considerably higher
than the recommended maximum error value of +/-
5%, as declared by BS EN 61685:2002-IEC 61685:2001,
a current International standard (stability date: 2020)
for the development of a flow Doppler test object [19].
It is this context that is the motivation for the
development of a novel phantom, designed to
produce complex flows that are stable, predictable,
controllable and reproducible. The particular flow
chosen as the basis for the design is the vortex ring.
This paper describes the salient features of a
prototype ring vortex phantom providing necessary
details for construction in a form suited to calibration
of Ultrasound flow imaging systems. With its uniquely
controllable characteristics, the ring vortex is a
suitable flow reference for a flow test object.

2. Materials and Methods
The complex flow phantom design described here
uses a piston to propel a slug of fluid along a cylindrical
channel, through an orifice port that connects to an
open tank of fluid (containing water, Blood-Mimicking
Fluid etc.) where propagation of the ring may be
observed and imaged. A vortex ring naturally forms
when a column of fluid is pushed through an orifice
into a neighbouring expansive fluid environment. The
        
face, forming a toroidal core of vorticity that may
ultimately detach (at high enough Reynolds number)
and propagates along the axisymmetric axis of the
ring. The whole process is managed by computer
control of piston displacement that delivers the fluid
slug with a precisely specified velocity/displacement
profile. Each component of the phantom is presented
below followed by description of the functioning unit
as a whole.
2.1 Imaging Tank
The imaging tank (Figure 1) is the principal component
of the phantom since it is the environment in which
the ring vortex propagates; it is the volume used to
image the generated flow. The tank is manufactured
from clear poly(methyl methacrylate) (PMMA) which
rests on four small screw feet. These are adjustable to
allow levelling of the system via two spirit levels
placed on opposite walls of the tank box. The tank has
internal dimensions of 15 cm (W) x 35 cm (L) x 16.5 cm
(H) (Figure 1), chosen to be sufficiently large that the
walls do not influence ring vortex propagation.
Several reference markers and a ruler are laser-cut
into the surfaces of the tank walls to support
Figure 1. CAD drawing of the ring vortex tank.

Figure 2. CAD drawing of the piston cylinder system.
Figure 3. CAD drawing of the attachable/detachable orifices.

Frequently asked questions

Q1. What have the authors contributed in "A complex flow phantom for medical imaging: ring vortex phantom design and technical specification" ?

Quality Assurance of technologies that image such flows is challenging but essential, and to this end, a novel, cost-effective, portable, complex flow phantom is proposed and the design specifications are provided. 

Q2. What are the future works in "A complex flow phantom for medical imaging: ring vortex phantom design and technical specification" ?

It is worth noting that their analysis has focussed on bulk flow characteristics ( eg. translational vortex ring speed Vtrans ) but for completeness, further work is needed to assess the micro-flow environment in addition to the macro-flow characteristics described here. Future work will extend the assessment presented here to include comparative studies between medical imaging modalities ( Ultrasound, CT, and with some adaptation MRI ) and optical modalities ( Laser-PIV, Laser-diode ) to further assess reliability, long-term stability and detailed flow performance. Currently, four identical phantom systems have been manufactured and are currently being evaluated in the United Kingdom and France within both research and clinical environments, in order to identify potential improvements to the design.