Lung Imaging in the Mouse with SWIFT Curt Corum, Djaudat Idiyatullin, Steen Moeller, Ryan Chamberlain, Deepali Sachdev, and Mike Garwood Center for Magnetic.

Lung Imaging in the Mouse with SWIFT

Curt Corum, Djaudat Idiyatullin,Steen Moeller, Ryan Chamberlain,Deepali Sachdev, and Mike Garwood

Center for Magnetic Resonance Research and Masonic Cancer Center, School of MedicineUniversity of Minnesota, Minneapolis, Minnesota, USA

#204

5/4/2010 ISMRM 2010 #204 Lung Imaging in the Mouse with SWIFT

2

Declaration of Conflict of Interest or Relationship

Speaker Name: Curt Corum

I have the following conflict(s) of interest to disclose with regard to the subject matter of this presentation:

Dr. Corum is entitled to sales royalty from Steady State Imaging, which isdeveloping products related to the research described in this presentation.


3

Motivation

Lung and especially lung parenchyma is difficult to image with MRI

Low proton density, susceptibility, T2, T

2*, cardiac and

respiratory motion all contribute At higher fields intrinsic (air to tissue and pathology)

and extrinsic (contrast injection) suseptibility effects become stronger

With T2

* sensitive sequences these cause signal dropout and phase effects due to the local field changes


4

Lots of interest and great work... UTE approachHuman and AnimalBergin, Noll et al.1992Shattuck, Gewalt et al.1997

GRE approachHuman and AnimalMuch great work...

FID (BLAST, RUFIS)Animal ImagingKuethe et al. and others...


5

AQ

Gro

RF

UTE excite → rampBLAST andRUFIS ramp → excite

SWIFTGRE

AQ

Gro

RF

HSn

AQ

Gro

RF

Sequences for capturing short T2 signals


6

T2

* of Parenchyma

Mouse LungGRE TE = 2msRespiratory gated

Clip level adjustment


7

SWIFT Sequence

SWeep Imaging with Fourier Transform Acquisition occurs in the gaps of a frequency swept (usually

HSn) pulse Excitation and Acquisition nearly simultaneous

“dead time” ~2 µs No time for slice selection or phase encoding, is most naturally a

readout only, interleaved preparations are possible Aquired data are FIDs (after correlation with the RF pulse

shape) Operates in 2d projection mode or 3d image mode with radial

FID sampling scheme and gridding reconstruction Most similar to BLAST, RUFIS, etc. in concept


8

SWIFT Sequence

Idiyatullin, D.; Corum, C.; Park, J. Y. & Garwood, M., Fast and quiet MRI using a swept radiofrequency., J Magn Reson, 2006, 181, 342-349


9

Mouse Model of Lung Cancer

7-9 week old athymic mice

Lung seeking metastatic variant of MDA-MB-231 breast cancer cells injected into the tail vein.

D. Sachdev

Isoflurane O2 N

2O anesthesia

N=4 Healthy Animals and

N=2 Tumor animals so far...


10

MRI parameters

• SWIFT TR 2.2 ms dead time 2 µs

• 125 kHz acquisition bandwidth

• 3072*16*2 = 96,304 unique FID views

• 256 complex points per view

• Prospective respiratory trigger and gating, 40-50% duty

• 100-200 views per trigger (typically 128)

• Reconstruction to 256^3 3D image

• ~0.12 mm nominal resolution

• 8-9 min (x4) for 32-36 min total imaging time per animal


11

Healthy Mouse

Respiratory pillow gated, 125kHz SWIFT9.4 T Varian animal system Combination of 4x 8 min scans (32 min)Quadrature surface coil on backAnimal is prone, ~3x3x4 cm fov

Intensity flattening appliedReformatted to bronchial VNative intensity


12

Tumor Mouse

Same parametersLeft lung has collapsed

Intensity flattening appliedReformatted to bronchial VNative intensity


13

Tumor Montage


14

Excised Lung

D. Sachdev


15

Discussion

Preserves signal Lung Parenchyma and signal next to susceptibility boudaries (bronchii to parenchyma, vessels to parenchyma, tumor margin to parenchyma, etc...)

Avoidance of gradient ramp sampling required for UTE sequences

Avoidance of acuostic noise and eddy currents Low peak RF power compared to BLAST or RUFIS

(radial FID aquisition sequences), T1 weighting

obtainableScalable to Human Imaging


16

Future Work

Histological and Pathological verificationnormal tissue, tumors, inflammation

Scalable to human imaging (no peak power or gradient performance limitations)

Contrast and DCE


17

Support

NIH BTRR 5P41RR008079-17CMRR Center Grant, Core 3, PI Mike Garwood

Masonic Cancer Center (startup), PI Deepali Sachdev 1R21CA139688-01, PI Curt Corum

IMPROVED BREAST DCE MRI WITH SWIFT MN MED FDN/3932-9227-09, PI Curt Corum

MRI Utilizing SWIFT to Detect Breast Calcifications,Minnesota Medical Foundation


18

Upcoming SWIFT presentations

• Theoretical Sensitivities of SWIFT and the Ideal Sequence (Delta Pulse-Acquire) for Ultra-Short T2Tuesday 4 May, 13:30 - 15:30 #2975

• Detection of Short T2 Component in Brain by SWIFTTuesday 4 May, 13:30 - 15:30 #2980

• In Vivo SWIFT Imaging of SPIO Labeled Stem Cells Grafted in the HeartTuesday 4 May, 14:30 - 15:00 #3747, Computer 40

• Measurement of T1 Relaxation Time in Articular Cartilage Using SWIFTWednesday 5 May, 13:30 - 15:30 #841

• Dipole Matched Filter with SWIFTWednesday 5 May, 14:30 - 15:00 #5113, Computer 126

• SWIFT Versus X-Ray In Dental ImagingThursday 6 May ,12:18 - 12:30 Room A5 #543


19

Bonus: Detailed SWIFT Timing


20

Bonus: Dipole Matched Filter

Bo

Unfiltered Magnitude Image Re ImageAfter dipole matched filter Fin k-space

Mag ImageAfter dipole matched filter Fin k-space

62kHz SWIFT at 4 T2 o'clock position has MRI compatible catheter tip

F=3cos2 θ k −1

2E-Poster #5113Thurs


21

Bonus: Teeth

With Don Nixdorf, DDS, Hari Prasad, et al.See talk #543 Thurs 05/06, 12:18 PM, room A5

Lung Imaging in the Mouse with SWIFT Curt Corum, Djaudat Idiyatullin, Steen Moeller, Ryan Chamberlain, Deepali Sachdev, and Mike Garwood Center for Magnetic.

Documents