MR Research Facility, Wayne State University, Detroit, Michigan nmrimaging@aol.com

THE ROLE OF SWI IN DETECTING

ABNORMAL OXYGEN SATURATION AND

MICROBLEEDS IN STROKE

E. Mark Haacke, PhD

Disclosures

Dr. Haacke is affiliated with:

Wayne State University, McMaster University,

University of Saskatoon, Loma Linda University,

Case Western Reserve University, East China

Normal University, and Northeast University.

The MRI Institute for Biomedical Research

MR Innovations, Inc.

Clinical applications of SWI and SWIM

See www.swim-mri.com

Our work in Detroit at Wayne State University

See www.mrc.wayne.edu

David Utriainen, for image analysis

Sean Sethi, MS, for image analysis

Meng Li, MS, for perfusion analysis

Jaladhar Neelavalli, PhD for SWIM support

Zhifeng Kou, PhD for TBI data

Sagar Buch, PhD for oxygen saturation measures

Saifeng Liu, PhD for advances in SWIM processing

Xia Shuang, MD for SWI stroke data

Meiyun Wang, MD for DE SWI data

And for a tremendous effort on STAGE Yongsheng Chen

Acknowledgements

SWI

Haacke EM et al. Susceptibility weighted imaging. Magnetic Resonance in

Medicine, 52: 612; 2004.

Susceptibility Weighted Imaging from a single

echo sequence with a long TE of 20ms at 3T

Enhances the presence of ferritin, hemosiderin and deoxyhemoglobin

Exquisite images showing venous structures and CMBs (cerebral microbleeds)

Using the venous signal, one can more easily see changes in deoxyhemoglobin in stroke

Cerebral amyloid angiopathy

50µ objects can manifest as 1mm3 objects

MRI scan date: 2013.01.04

MRI scan date: 2013.01.11

Two scans from the same stroke patient

MTT

SWI

SWIMTT

Images courtesy of Luo Yu MD, The Branch of Shanghai First Hospital

Venous

Thrombosis:

before

treatment

and after

thrombolysis

Guangbin Wang M.D.

Shandong Medical Imaging

Research Institute

T1 T2

PRE POST

SWI SWIM

SWI enhances the presence of ferritin, hemosiderin and deoxy-Hb; provides exquisite images to diagnose microbleeds and abnormal oxygenation levels

SWIM allows direct quantitative analysis for abnormal iron in brain tissues and deoxy-Hb levels

SWI VS. SWIM

Minimum intensity projection of SWI over 28 slices

SWI PRE

CAFFEINE

SWI POST

CAFFEINE

SWI as a high resolution BOLD imaging method

200mg caffeine pills (a, d)

or 1000mg diamox IV

injection (c, f).

Compared to the control

condition (b, e),

significant oxygen

saturation changes are

observed post-challenge

on veins throughout the

brain.

SWI

SWIM

Imaging veins and blood products using SWI and SWIM:

Challenging the neurovascular system

Caffeine: flow change = − 30% ± 5% and ΔY = − 0.09 ± 0.02

Diamox: flow change = +31% ± 3% and ΔY = +0.10 ± 0.01

a

b

c

MIPs of SWIM data over 8mm. The data where collected at 3T with TE=14.3ms and a voxel size 0.5x0.5x0.5mm3.

Future Directions:

OXYGEN EXTRACTION FRACTION

Green - deoxyhemoglobin levels in the veins

Blue represents iron in the basal ganglia and midbrain

Whole brain iron oxygen extraction fraction

Materials and Methods

Subjects 26 patients with cerebral hemisphere ischemic stroke (screened

from 78 cases) 30 healthy volunteers age and sex matched

Criteria All patients had SWI, DWI and MRA DWI high signal、ADC low signal, suggests acute ischemic stroke Only unilateral hemisphere involvement (contralateral served as

control) Cortical veins of bilateral hemispheres are easy to visualize

Venous oxygen saturation evaluation

• Stroke case

Decreased oxygen saturation in asymmetrically prominent

cortical veins in patients with cerebral ischemic stroke.Xia S, Utriainen D, Haacke EM, et al.

Magn Reson Imaging. 2014 Dec; 32(10):1272-6.

Venous oxygen saturation evaluation

• Stroke cases

Decreased oxygen saturation in asymmetrically prominent

cortical veins in patients with cerebral ischemic stroke.Xia S, Utriainen D, Haacke EM, et al.

Magn Reson Imaging. 2014 Dec; 32(10):1272-6.

Venous oxygen saturation

evaluation

• Stroke case

High resolution MR angiography:

Where are the cerebral vessels?

11) Salamon, G., 1971. Atlas of the arteries of the human brain. Sandoz, Paris.

250μ x 250μ x 500μ MRA using a contrast agent, 15 minute scan

Visualizing Small Arteries

Small arteries around 250 microns and possibly even smaller

are becoming visible (see thalamic arteries)

MRA short echo SWI

SWI only veins At 3T, veins are more naturally

suppressed because they have

T2* = 25ms while arteries have

a T2* closer to 70-80ms. Images

acquired with isotropic in-plane

resolution of 0.5mm x 0.5mm

in-plane and TH = 1mm.

Images courtesy of Yongquan Ye, PhD

Simultaneous MRV and

MRA using a double

echo interleaved SWI

rephased/dephased

sequence

Simultaneous MRV and

MRA using a double

echo interleaved SWI

rephased/dephased

sequence

MRA short echo SWI RP-DP MRA

SWI only veins

With no contrast agent

the interleaved double

echo sequence offers

excellent background

suppression to give a

superb MRA and SWI.

Images courtesy of Yongquan Ye, PhD

RP-DP MRA

Imaging veins and arteries using double echo SWI

Thrombus dominates the SWI

image (TE = 7.5ms)

Note the asymmetrically

prominent cortical veins

Thrombus dominates the

SWIM image (TE = 7.5ms)

First echo MIP

Second echo (17.5ms) tSWI

First echo magnitude

First echo SWI phase image

showing the MCA wall

Images courtesy of Meiyun Wang, MD, Henan Provincial People’s Hospital

Stroke case for a young woman in her mid 30s

FLAIR PWI (1st) SWI mIP MRA

CBV CBF MTT TTP

STAGE short protocol = 6 minutes total scan time*

Welcome to STAGE Imaging: Whole brain imaging in just a few minutes!

From the lab of E. Mark Haacke, PhD

T1 SWI SWIM

T1 map R2* map MRA

T2

ADC map

Add 3 more minutes…

6 minutes scan time based on 0.67x1.34x2.68 resolution

Imaging the whole brain within 5 minutes with STAGE A provides

not only the usual conventional images such as:

• A) T1W, PDW, and T2* weighted images but also

• B) a T1 MAP, PD MAP, R2* MAP and over and above this

• C) SWI, MRA and SWIM images and finally the potential for

• D) Tissue segmentation for CSF, GM, WM, arteries and veins

Strategically Acquired

Gradient Echo Imaging or

STAGE Imaging

The rapid STAGE protocol consists of collecting the data with a

resolution of:

0.67 x 1.3 x 2.7mm3 and 48 (64) partitions.

The basic protocol consists of the follow SWI scans:

• B1: TE=7.5ms/18.5ms, FA=6o, 2 (2.5) minutes

• B2: TE =7.5ms/17.5ms, FA = 24o, 2 (2.5) minutes

• B3: TE=5ms rephased/17.5ms dephased, FA=24o, 4 minutes

• Data Processing done by SPIN software from MR Innovations.

STAGE 2 A and B Imaging

The rapid STAGE protocol can be augmented by adding:

• One of FLAIR or SPIN ECHO 2D or 3D scans in 2 minutes

• DWI in 2 minutes

and for stroke cases, if thought appropriate

• PWI in 2 minutes

Data Processing done by SPIN software from MR Innovations.

Augmented STAGE Imaging

STAGE for an old stroke case

SWI, tSWI and SWIM images from the FA=6o, TE=17ms data.

These images represent a mIP/MIP of 11 slices

The hypointense area in the left brain on SWI which has a positive susceptibility on

SWIM indicates a hemorrhage or an abnormal vein in this area.

SWI SWIM True SWI

Dr. Mark Haacke and Yongsheng Chen

Sample Images from a thumb to opposed

finger tapping event as seen for the RH.

Original magnitude images from TE=20ms. Signal in the magnitude = 280 units and in

the difference image 20 units. A standard error of 1.7 implies an SNR of .

So we see a 7% change in the PMC with an SNR of 165:1. The difference error is 2.4

giving the final result of an activation of 7% +/- 1%.

Future Directions

Imaging atherosclerosis for stroke patients in

The carotid arteries

The MCAs

Imaging vulnerable plaque

Imaging the entire vasculature with SWI by using a USPIO such as Ferumoxytol and the MICRO imaging protocol

Carotid vessel wall plaque, TE = 5ms

Imaging vessel wall using SWI and SWIM

TE = 15.6ms

0.5mm x 0.5mm x 1mm

sagittal acquisition

8 minutes without parallel

imaging

2008 time frame

0.37 radian phase shift

Carotid artery SWI and SWIM

This may be a case of vulnerable plaque. Here you can see what appears to be a small thrombus on the inside wall.

If that is the case, it could break off and become an embolus causing a stroke.

SWIM validates it is iron and therefore likely blood.

magnitude SWIM

MRA

phase

phaseFlow compensated, TE

= 5ms with an inplane

resolution of

0.5mm x 0.5mm and 64

slices 2mm thick

Scan time 4min, 36sec

Images courtesy of Hyun Seok Choi MD and Eo-Jin Hwang from Seoul, Korea

A B C

MICRO

“Microvascular In-vivo Contrast Revealed Origins”

Ferumoxytol enhanced MRAV

Image courtesy of Yulin Ge MD, NYU.

The process of separating arteries from

veins from pre/post Ferumoxytol images.

MICRO data acquired at 7T with TE = 8ms and a

resolution of 100μm x 200μm x 1.25mm.

(A) MIP of pre-contrast original magnitude image

(1st echo) showing the arterial signal (red).

(B) Post 4mg/kg Ferumoxytol MICRO image.

(C) Basal ganglia arteries from the cadaver brain

work of Georges Salamon (1971).

(D) mIP of pre-contrast SWI shows veins only.

(E) Post-Ferumoxytol SWI (2mg/kg) showing both

veins and arteries including small arterioles

(red arrowheads) and venules (blue

arrowheads).

(F) Post-contrast QSM showing higher

susceptibility values in veins than in arteries.

Our goal is to track both arteries and veins, and use

this to study microvascular disease at 3T.

Arteries, arterioles and venules are now visible

thanks to the susceptibility contrast and blooming

effect from Feuromxytol.

A B C

D E F

Conclusions

SWI offers new possibilities for studying stroke

SWI/PWI mismatch suggests collateral flow and potentially better outcomes

SWI detection of CMBs may impact treatment

Stroke patients should be imaged at the acute stage and again before discharge a week later in order to better understand patient recovery

STAGE imaging can be used to: image stroke in less than 10 minutes, image thalamic nuclei and possibly in the future for fMRI

Microvascular imaging is possible with USPIO and SWI to image vessels down to 50-100μ in diameter