Notes
Slide Show
Outline
1
Visualization of Hemodynamics in the Field of Neuroradiology using Time-Resolved Three-Dimensional Phase-Contrast MR Imaging
  • Haruo Isoda1, Masaya Hirano2, Shuhei Yamashita1, Shoichi Inagawa1, Takashi Kosugi3, Marcus T. Alley4,
  • Michael Markl4, Norbert J. Pelc4 and Harumi Sakahara1
2
Contents
  • Purpose
  • Basic Principle of Cine Phase Contrast (PC) MR Imaging
  • Basic Principle of Time-Resolved Three-Dimensional Phase-Contrast MR Imaging (4D-Flow)
  • Parameters and Postprocessing of 4D Flow Imaging
  • Images Analyzed by Visualization Software
  • In Vivo Hemodynamic Analysis of a Carotid Bifurcation
  • In Vivo Hemodynamic Analysis of Normal Intracranial Arteries
  • In Vitro Hemodynamic Analysis of a Silicon MCA Bifurcation Aneurysm Model
  • In Vivo Hemodynamic Analysis of an A-com Aneurysm
  • Discussion
  • References
3
Purpose
  • Hemodynamics affects the development and growth of vascular lesions, such as atherosclerotic plaques and aneurysms (1).  Being able to predict the occurrence of vascular lesions and their growth, based on hemodynamics, would enable us to judge the outcome of the patient.  This information would be beneficial for the prevention of disease and the planning of treatment for each patient.  To date, two-dimensional cine phase contrast MR imaging technique and other imaging techniques have not been adequate for this purpose.
  • Recently a new cine phase contrast MR imaging technique named Time-Resolved Three-Dimensional Phase-Contrast MR Imaging (4D-Flow) (2, 3) has been developed.  It provides us with in vivo four-dimensional hemodynamic information including space and time.  The purpose of this exhibit is to introduce the 4D-Flow technique and to show its usefulness in the field of Neuroradiology.
4
Basic Principle of Cine Phase Contrast (PC) MR Imaging
  • The upper part demonstrates a bipolar gradient.  Horizontal axis is time.
  • The lower part shows the phase shifts of stationary tissue, fast flow and slow flow.  Three lines demonstrate fast flow, slow flow and stationary tissue.  The longitudinal axis is phase shift created by bipolar gradient.  Horizontal axis is time.  After the end of the exposure of a bipolar gradient, phase shift of stationary tissue created by bipolar gradient is zero.  Phase shifts of other flows created by bipolar gradient are shown in the figure.  There is a correlation between phase shift created by bipolar gradient and velocity.
5
Basic Principle of Time-Resolved Three-Dimensional Phase-Contrast MR Imaging (4D-Flow)
  • This imaging technique was reported by Markl M, et al.(2)
  • The 4D-Flow technique is based on radio frequency-spoiled gradient-echo sequence and velocity encoding is performed along all three spatial directions.  Four dimensional (4D) data including time dimension are obtained.  Measurements are retrospectively gated to the electrocardiogram and CINE series of three-dimensional (3D) data sets are generated.
  • In this 4D-Flow imaging sequence, 4 sets of longitudinal bands represent TR intervals for acquisition of magnitude images and for encoding velocities in 3 directions. Total number of phase-encoding steps is Nky and total number of slice-encoding steps is Nkz.  nkz is the number of lines filled in k-space at once. When nkz is 4, temporal resolution is 16TR.  The total acquisition time is given by Nky・Nkz・TECG/nkz, with TECG = average cardiac period.
6
Basic Principle of Time-Resolved Three-Dimensional Phase-Contrast MR Imaging (4D-Flow)
  • Twenty phases of 3D data set with magnitude information and 3 directional velocity information for one cardiac cycle are obtained using 4D-Flow. The images encoded for x-y-z directions are corrected by magnitude image and 3D velocity data for each 3D voxel is generated. The data including velocity components in x, y, z direction and time provide us with 4D flow information.
7
4D–Flow Imaging
  • Scanner
    • 1.5T MR scanner (Signa Infinity Twinspeed with Excite XI, GE Medical systems, Milwaukee, WI)
      • Carotid, 8 channel neurovascular coil; Intracranial arteries, 8 channel head coil; in vitro study, QD head coil
  • Imaging parameter
    • In vivo study
      • Carotid
        • TR/TE/NEX = 4.8/1.6/1, FA = 15degrees, FOV = 200x200x48mm, Matrix = 192x192x16, voxel size = 1.04x1.04x3mm, Velocity encoding (VENC) = 60cm/s, 20 phases during one cardiac cycle, imaging time=15-20min, sagittal plane
      • Intracranial arteries
        • TR/TE/NEX = 5.4/2.3/1, FA = 15degrees, FOV = 160x160x40mm, Matrix = 160x160x40, voxel size = 1x1x1mm, VENC = 60cm/s, 20 phases during one cardiac cycle, imaging time=20-40min, transverse plane
    • In vitro study
      • TR/TE/NEX = 5.8/2.1/1, FA = 15degree, FOV = 140x140x108mm, Matrix = 160x160x36, voxel size = 0.88x0.88x3mm, VENC = 20cm/s, 20 phases during one cardiac cycle, imaging time = 48min, transverse plane
  • Postprocessing
    • Time-resolved images of 3D streamlines, 3D particle traces, 3D pathlines and two-dimensional (2D) velocity vector fields on arbitrary planes were calculated from 4D data sets by flow visualization software (EnSight).
8
Images Analyzed by Visualization Software
  • 2D velocity vector field
    • arbitrary 2D planes extracted from the 3D imaging volume; 3 directional velocities are displayed over time as color-coded vector fields (3)
      • A left Carotid Bifurcation
      • A left IC-PC Portion
      • A Silicon MCA Aneurysm Model
  • 3D streamlines
    • integrated traces along instantaneous velocity vector field, color coded according to the local velocity magnitude (3)
      • A Circle of Willis
      • A Silicon MCA Aneurysm Model
  • 3D particle trace
    • path that a massless particles would follow if placed in a time-varying vector field, color-coded according to the local velocity magnitude (3)
      • A left Carotid Bifurcation
      • A left Carotid Siphon
  • 3D pathlines
    • integration of particle traces
      • A left Carotid Siphon
      • An A-com Aneurysm
9
In Vivo Hemodynamic Analysis of the Carotid Bifurcation
10
A left Lateral View of 2D Velocity Vector Field
on a Plane Traversing the left Carotid Bifurcation
for a 47-year-old Normal Male Volunteer
  • A.  A MR angiogram of the left carotid bifurcation shows the shape of the vascular structure.
  • B.  2D velocity vector field traversing the left carotid bulb clearly shows that the flow separation (arrows) is clearly noted at the posterior aspect of the carotid bulb.  The vectors are color-coded according to the velocity shown at the legend.  Unit of the legend is m/sec.
11
3D Particle Traces of the left Carotid Bifurcation
for a 47-year-old Normal Male Volunteer
  • 3D particle trace of the left carotid bulb clearly shows the presence of the flow separation or recirculation at the posterior aspect of the carotid bulb.  The vectors are color-coded according to the velocity shown at the legend.  Unit of the legend is m/sec
12
In Vivo Hemodynamic Analysis of Normal Intracranial Arteries
13
3D Streamlines around the Circle of Willis of a 23-year-old Normal Male Volunteer
  • Craniocaudal views of 3D streamlines around the Willis’ circle clearly demonstrate the Willis’ circle, bilateral M2 segments of the middle cerebral artery, bilateral A2 segments of the anterior cerebral artery and bilateral P3 segments of the posterior cerebral artery.  These images can be displayed as CINE images.  Streamlines were generated from the bilateral internal carotid arteries and the basilar artery. Velocities are expressed as color scale.  Unit of the legend is m/sec.
14
3D Particle Traces of the left Carotid Siphon of
a 22-year-old Normal Male Volunteer
  • Lateral views of 3D particle traces show parabolic velocity profile in the internal carotid artery.  Note the laminar flow.
15
3D Pathlines of the left Carotid Siphon of
a 22-year-old Normal Male Volunteer
  • Pathlines in which traced particles are connected show that pathlines are parallel at the straight portion of the vessel and that they are helical in the region of vessel curvature.  Particles are generated from the C5 segment of the left internal carotid artery.  Color of particles and lines represent velocity.  Unit of the legend is m/sec.
16
A left Lateral View of 2D Velocity Vector Field
on a Plane Traversing the left IC-PC Portion
for a 23-year-old Normal Male Volunteer
  • A.  A magnitude image traversing the left IC-PC portion shows the shape of vascular structure.
  • B.  2D velocity vector fields clearly show flow vectors around the IC-PC portion.  Flow vectors just adjacent to the vascular wall are clearly depicted. The vectors are color-coded according to the velocity shown in the legend.  Unit of the legend is m/sec.
17
In Vitro Hemodynamic Analysis of a Silicon MCA Bifurcation Aneurysm Model
18
Silicon Aneurysm Model
  • 1. Phantom
  • 1) A silicon intracranial aneurysm model 3 times actual size
  • Rotational digital subtraction angiography (Fig. 1) was performed for a 67-year-old female with a right middle cerebral aneurysm. The greatest dimension of this intracranial aneurysm was 8mm.  The aspect ratio (defined as depth/neck width) of the aneurysm (8) was 4.8. Based on the DICOM data sets of the angiogram, a realistic silicon model, 3 times actual size, containing the lumen of the original vessels was then constructed (Fig. 2).  Therefore, the size of the intracranial aneurysm in our model was 24mm.
  • 2) Phantom circuit (Fig. 3)
  • A phantom circuit was a closed circuit with a reservoir tank (A), a computer controller (B), a pulsatile pump (CardioFlow 1000MR, Shelly Medical Imaging Technologies company, Toronto, Canada) (C), the silicon model (D) and simulation vessel (E).  We ran a 53% weight glycerin solution (T1 value, 1005msec; T2 value, 86msec) in pulsatile flow (F) with the pump.  One cardiac cycle was 2 seconds and the maximum systolic velocity was 84cm/s.  Nondimensional parameters, Reynolds number (811) and Womersley number (3.26) were similar to those of human blood flow.  Blood flow conditions in the patient were accurately reproduced in our silicon model.
19
Anteroposterior Views of Rotational Angiogram, 3D Streamlines and a 2D Vector Field traversing M1 Segment and two M2 Segments of the right MCA for the Silicon MCA Aneurysm
  • Surface rendering image (A) obtained from rotational digital subtraction angiogram of right internal carotid artery reveals an aneurysm in the bifurcation of MCA.  M1, M2 segments are well identified in frontal view (B) of 3D streamlines in systolic phase.  Counter clockwise vortex flow in the aneurysm is identified on 2D velocity vector field traversing the M1 segment and two M2 segments of right MCA (C).  Numbers and colors correspond with flow rate in the legend at the right side in the figure regarding 3D streamlines and 2D velocity vector fields.  Units in m/s.  Flow rate in the aneurysm was less than 15cm/s.
20
2D Velocity Vector Fields of the Silicon MCA Aneurysm Model
  • Kinematic display of 2D velocity vector fields shows that counter clockwise vortex flow in the aneurysm is clearly identified on a plane traversing the M1 segment and two M2 segments of right MCA.  In addition, 2D velocity vector fields clearly demonstrate flow velocity near the vascular wall.
21
Left Anterior Oblique Views of a Rotational Angiogram
and 3D Streamlines of the Silicon MCA Aneurysm
  • A.  A rotational angiogram. Arrows indicate the bleb.
  • B.  Blood flow from the M1 segment of MCA strikes the posterior wall of the intracranial aneurysm.  Blood flow flows along the wall of the intracranial aneurysm (the highest flow rate, 15cm/s) and diverges into the two M2 segments near the inlet of the aneurysm.  Curved arrow indicates flow.
  • C.  A helical flow at the left aspect of the aneurysm is seen along the aneurysmal wall.  Where the helical flow reverses direction is coincident with the bleb (shown by arrows in A).  Flow rate of the bleb is low (less than 5cm/s).  Spiral arrow indicates flow.
  • D.  Helical flow continues to flow into the center of the aneurysm from the bleb.  Flow rate is less than 5cm/s.  Spiral arrow indicates flow.
22
Posterior-anterior Views of a Rotational Angiogram and 3D Streamlines of the Silicon MCA Aneurysm
  • A.  A rotational angiogram. Arrows indicate irregular surface of the aneurysm.
  • B.  Blood flow from the M1 segment of MCA strikes the posterior wall of the intracranial aneurysm.  Blood flow flows along the wall of the intracranial aneurysm (the highest flow rate, 15cm/s) and diverges into the two M2 segments near the inlet of the aneurysm.  Curved arrow indicates flow.
  • C.  A helical flow at the left aspect of the aneurysm is seen along the aneurysmal wall.  Spiral arrow indicates flow.  This helical flow with low velocity corresponds with the irregular surface of the aneurysmal wall (shown by arrows in A).
  • D.  Helical flow continues to flow into the center of the aneurysm from the bleb.  Flow rate is less than 5cm/s.  Spiral arrow indicates flow.
23
Kinematic display of 3D streamlines rotating along the head-foot axis during systolic phase of the Silicon MCA Aneurysm
  • 1) Blood flow from the M1 segment of MCA strikes the posterior wall of the intracranial aneurysm.
  • 2) Blood flow flows along the wall of the intracranial aneurysm (the highest flow rate, 15cm/s) and diverges into the two M2 segments near the inlet of the aneurysm.
  • 3) A helical flow at the left aspect of the aneurysm is seen along the aneurysmal wall. This helical flow with low velocity corresponds with the irregular surface of the wall.
  • 4) Where the helical flow reverses direction is coincident with the bleb.  Flow rate of the bleb is low (less than 5cm/s).
  • 5) Helical flow continues to flow into the center of the aneurysm from the bleb.  Flow rate is less than 5cm/s.
24
Kinematic display of time-resolved 3D streamlines during one cardiac cycle in anteroposterior direction of the Silicon MCA Aneurysm
  • Blood flow changes are observed in each phase.
25
In vivo Hemodynamic Analysis of an A-com Aneurysm
26
MR angiogram and 3D pathlines of a 59-year-old male patient with an A-com Aneurysm 5mm in size
27
Kinematic display of 3D pathlines of an A-com Aneurysm 5mm in size
  • Vortex flow in the A-com aneurysm is clearly shown by 3D pathlines.  Flow in the A-com aneurysm mainly comes from the right A1 segment.  Pathlines were generated from the distal part of each A1 segment of the right and left anterior cerebral arteries.
28
Discussion
  • Normal Vascular Hemodynamics
  • Hemodynamics and the Location of Vascular Diseases
  • Wall Shear Stress
  • Aneurysm
  • The Present Limitation of 4D Flow
  • The Future Utility of 4D Flow in the Field of Neuroradiology
29
Normal Vascular Hemodynamics 1
  • Reynolds number (Re) = ρUD/μ
        • ρ, density
        • U, velocity in the vessel
        • D, vessel diameter
        • μ, viscosity

  • Re is the ratio of inertial to viscosity forces in a fluid.
  • Re in the intracranial arteries < 2000
  •                     → Laminar flow,      Parabolic flow profile


  • Large carotid bulb
    • Flow separation zone and reverse flow are present at the posterior aspect of a large carotid bulb (4).
30
Normal Vascular Hemodynamics 2
  • Straight portion of the artery (5, 6)
    • Laminar flow
    • Parabolic flow profile
    • The fluid slipstreams are parallel to the vessel sidewall.
  • In the region of vessel curvature (5, 6)
    • The fluid slipstreams are helical.
    • The blood seemes to flow in the greater curvature.
    • The recirculation zone is present along the lesser curvature.
  • Bifurcation (5, 6)
    • A lower shear region forms at the lateral wall of the daughter limb.
    • If the bifurcation angle is large enough or the velocity is high enough, an actual separation or recirculation zone may form.
31
Hemodynamics and the Location of Vascular Diseases
  • Hemodynamics is suspected to be one of the main causes of vascular diseases such as atherosclerosis and intracranial aneurysms, because they usually develop near the vascular bifurcation (1).
  • Atherosclerosis
    • Posterior portion of the carotid bulb
    • Low shear stress
  • Intracranial aneurysm
    • Lateral aspect of the bend just after bifurcation such as the internal carotid artery posterior communicating artery bifurcation, the anterior communicating artery bifurcation or the middle cerebral artery bifurcation
    • High shear stress
32
Wall Shear Stress 1
33
Wall Shear Stress 2
  • Vascular wall shear stress of the appropriate strength (> 1.5N/m2) can maintain normal blood vessel function and low shear stress (< 0.4N/m2) causes atherosclerosis (1).
  • Vascular wall shear stress attracts attention as a cause of the development of intracranial aneurysms and their rupture (7-9).
  • Nitric oxide, derived from the endothelial cells, increases in the area with greater shear stress and causes vascular wall degeneration, vasodilatation and bleb dilatation (7).
  • When shear stress is low, apoptosis occurs, and danger of aneurysmal rupture rises (8, 9).
34
Aneurysm
  • Hemodynamics of aneurysms (10)
    • Blood flow enters into the intracranial aneurysm from the distal portion of the neck.  The blood flow collides directly with the aneurysmal wall.  Vortex flow occurs and then the blood flow flows out from proximal part of the neck peripheral to the distal central incoming jet.
  • Development of aneurysms
    • High wall shear stress is the cause of the development of intracranial  aneurysms (9).
  • Future possible site of the bleb
    • Two theories
      • The blebs showed higher instantaneous shear stress as compared with other sites in the aneurysms during cardiac cycle (7).
      • The shear stress of the intracranial aneurysmal wall was lower than that in the vascular wall and lower in the bleb than in other parts of the aneurysm (8, 9).
        • Blood flow in the bleb was slow, when the aspect ratio of the aneurysm (defined as depth/neck width) was higher than 1.6.
        • Lower wall shear stress may cause apoptosis of endothelial cells and rupture of the aneurysm.
35
The Present Limitation of 4D-Flow
  • The setting of VENC
    • We must choose proper VENC in order to prevent aliasing, because higher blood velocity will alias by producing a higher phase shift.
    • However, higher VENC is not suitable for visualizing slow flow in the intracranial aneurysms and the slow flow along the vascular wall.
  • The longer acquisition time
    • Image degradation due to motion might happen during this acquisition time.
    • The acquisition time for 4D-Flow is longer than other routine MR imaging however, we think that this technique can be applied for in vivo hemodynamic analysis for human intracranial aneurysms.
36
The Future Utility of 4D-Flow in the Field of Neuroradiology
  • Calculation of the wall shear stress
    • 4D-Flow clearly reveals hemodynamics near the vascular wall and the aneurysmal wall.  This suggests that intracranial aneurysmal wall shear stress can be obtained using this 4D-Flow technique in the future.
  • Tailor-made therapies and managements for each patient
    • If we could clarify the connection between the vascular lesions and in vivo human hemodynamics, in vivo hemodynamics measurement with MR imaging may be useful for prevention of vessel diseases, estimation of the outcome and deciding treatment plans for each patient.
  • Usage of hemodynamic information obtained by 4D-Flow as the boundary condition in order to improve the accuracy of results of CFD
37
Conclusion
  • The 4D-Flow technique is a promising non-invasive technique and provides us with time-resolved three dimensional in vivo hemodynamic information about intracranial and carotid arteries.
  • In the future we hope to be able to calculate wall shear stress based on data obtained by 4D-Flow.
  • The wall shear stress provided by this technique will hopefully enable us to predict the occurrence of atherosclerotic plaques and aneurysms and to predict the risk of aneurysmal rupture.
  • Furthermore, knowing this would aid us in preventing disease and making good therapeutic plans.
38
References
  • 1. Malek AM, Alper SL, Izumo S. Hemodynamic shear stress and its role in atherosclerosis. JAMA. 1999;282:2035-42.
  • 2. Markl M, Chan FP, Alley MT, Wedding KL, Draney MT, Elkins CJ, Parker DW, Wicker R, Taylor CA, Herfkens RJ, Pelc NJ. Time-resolved three-dimensional phase-contrast MRI. J Magn Reson Imaging. 2003;17:499-506.
  • 3. Markl M, Draney MT, Hope MD, Levin JM, Chan FP, Alley MT, Pelc NJ, Herfkens RJ. Time-resolved 3-dimensional velocity mapping in the thoracic aorta: visualization of 3-directional blood flow patterns in healthy volunteers and patients. J Comput Assist Tomogr. 2004;28:459-68.
  • 4. Motomiya M, Karino T. Flow patterns in the human carotid artery bifurcation. Stroke. 1984;15:50-6.
  • 5. Kerber CW, Heilman CB. Flow dynamics in the human carotid artery: I. Preliminary observations using a transparent elastic model. AJNR Am J Neuroradiol. 1992;13:173-80.
  • 6. Kerber CW, Liepsch D. Flow dynamics for radiologists. II. Practical considerations in the live human. AJNR Am J Neuroradiol. 1994;15:1076-86.
  • 7. Tateshima S, Murayama Y, Villablanca JP, Morino T, Nomura K, Tanishita K, Vinuela F. In vitro measurement of fluid-induced wall shear stress in unruptured cerebral aneurysms harboring blebs. Stroke. 2003;34:187-92
  • 8. Ujiie H, Tachibana H, Hiramatsu O, Hazel AL, Matsumoto T, Ogasawara Y, Nakajima H, Hori T, Takakura K, Kajiya F. Effects of size and shape (aspect ratio) on the hemodynamics of saccular aneurysms: a possible index for surgical treatment of intracranial aneurysms. Neurosurgery. 1999;45:119-29
  • 9. Shojima M, Oshima M, Takagi K, Torii R, Hayakawa M, Katada K, Morita A, Kirino T. Magnitude and role of wall shear stress on cerebral aneurysm: computational fluid dynamic study of 20 middle cerebral artery aneurysms. Stroke. 2004;35:2500-5
  • 10. Imbesi SG, Kerber CW. Analysis of slipstream flow in two ruptured intracranial cerebral aneurysms. AJNR Am J Neuroradiol. 1999;20:1703-5