Ultra High Field MRI Charles Dietz MD 2017

  • Slides: 75
Download presentation
Ultra High Field MRI Charles Dietz MD 2017

Ultra High Field MRI Charles Dietz MD 2017

2

2

Frequency directly related to Magnetic Field 3

Frequency directly related to Magnetic Field 3

4 Tesla (Installed in CMRR in ~1990)

4 Tesla (Installed in CMRR in ~1990)

4 TESLA (1991) MDEFT Barfuss et. al. NMR Biomed: 3(1)1990 (DATA from SIEMENS) SIEMENS,

4 TESLA (1991) MDEFT Barfuss et. al. NMR Biomed: 3(1)1990 (DATA from SIEMENS) SIEMENS, GE, and PHILIPS

BOLD f. MRI a 1991 4 Tesla CMRR R b Full Field Visual Stimulation

BOLD f. MRI a 1991 4 Tesla CMRR R b Full Field Visual Stimulation HEMI Field Visual Stimulation c Ogawa et al. Proc Natl Acad Sci USA (1992) 89, 5951 -

Functional Imaging with Magnetic Resonance Visual Stimulus 4 Tesla 2 1 3 Title Ogawa

Functional Imaging with Magnetic Resonance Visual Stimulus 4 Tesla 2 1 3 Title Ogawa et al. Proc Natl Acad Sci USA (1992) 89, 5951 -5955

Wei Chen et al. , U of Minnesota, CMRR, 4 T 4% R L

Wei Chen et al. , U of Minnesota, CMRR, 4 T 4% R L R 0. 2% L

Opaque anatomical Image In Grey scale SILENT WORD GENERATION

Opaque anatomical Image In Grey scale SILENT WORD GENERATION

DEVELOPMENT OF CMRR INSTRUMENTATION for HIGH MAGNETIC FIELD IMAGING and SPECTROSCOPY 4. 7 Tesla/40

DEVELOPMENT OF CMRR INSTRUMENTATION for HIGH MAGNETIC FIELD IMAGING and SPECTROSCOPY 4. 7 Tesla/40 cm (~1985) • 4 Tesla Human (1991) 9. 4 Tesla/31 cm (~ 1995) • 7 Tesla Human (~1999) 3 Tesla ~2002 • 9. 4 Tesla Human (~2005) 2 nd 7 T (2010) and 10. 5 T (2015) 16. 4 Tesla /26 cm bore (animal) (2010)

7 T MRI Department of Radiology Charles Dietz

7 T MRI Department of Radiology Charles Dietz

7 Tesla/90 cm bore ~1999

7 Tesla/90 cm bore ~1999

7 Tesla UHF Clinical MRI • Ultra Highfield (UHF) MRI at 7 T with

7 Tesla UHF Clinical MRI • Ultra Highfield (UHF) MRI at 7 T with increased S/N ratio • Spectroscopy • Higher Resolution – Improved Diagnostic Accuracy • Shorter Acquisition Time – Higher Clinical Throughput

Patellar Cartilage 7 T 3 T

Patellar Cartilage 7 T 3 T

3 T Knee MRI

3 T Knee MRI

7 T Knee MRI

7 T Knee MRI

Meniscal Tears 7 T 3 T

Meniscal Tears 7 T 3 T

3 TESLA SUBTLE FRACTURE

3 TESLA SUBTLE FRACTURE

7 TESLA SUBTLE FRACTURE

7 TESLA SUBTLE FRACTURE

DETAILED ANATOMY OF TIBIAL NERVE AT 7 T (CMRR) Knee imaging at 7 T

DETAILED ANATOMY OF TIBIAL NERVE AT 7 T (CMRR) Knee imaging at 7 T a) 16 -element transmit/receive for all knee images. b) Sagittal turbo spin echo (TSE) image (TR/TE = 4210/31 ms, resolution 0. 2 x 3, 39 slices, 1 average, echotrain length 7, 10 min. c) Axial turbo spin echo (TSE) image (TR/TE = 4680/27 ms, resolution 0. 3 x 2, 29 slices, 2 averages. Ellermann et al.

Quantitative cartilage assessment: T 1ρ -maps of articular cartilage Dog model of early osteoathritis

Quantitative cartilage assessment: T 1ρ -maps of articular cartilage Dog model of early osteoathritis at 7 Tesla (CMRR) T 1ρ Surgery preparation module nx 4 hyperbolic secant (HS 1)-pulses pulse duration: 5. 12 ms Control ms control surgery p (two-tail) medial 160± 30 210± 30 0. 01 imaging module in-plane resolution: 620 μm lateral 190± 50 270± 60 0. 7 slice thickness: 1. 5 mm matrix: 128× 128 T 1ρ of the cartilage in the severed joint is longer than in the control # segments: 4 TE/TR = 5. 1 ms/4 s Pepin SR et al: A comparative analysis of 7. 0 -Tesla magnetic resonance imaging and histology measurements of knee articular cartilage in a canine posterolateral knee injury model: a preliminary analysis. Am J Sports Med.

IMAGING AT HIGH AND ULTRAHIGH FIELDS IN THE HUMAN TORSO; THEN (1990; 4 TESLA)

IMAGING AT HIGH AND ULTRAHIGH FIELDS IN THE HUMAN TORSO; THEN (1990; 4 TESLA) 4 Tesla (Data from Siemens) Barfuss et al. NMR Biomed: 3(1)1990 NOW (7 TESLA) 7 Tesla Henry, TR et al. , Radiology 2011; 261(1): 199 -

IMAGING AT HIGH AND ULTRAHIGH FIELDS IN THE HUMAN TORSO; THEN (1990; 4 TESLA)

IMAGING AT HIGH AND ULTRAHIGH FIELDS IN THE HUMAN TORSO; THEN (1990; 4 TESLA) 4 Tesla (Data from Siemens) Barfuss et al. NMR Biomed: 3(1)1990 NOW (2009; 7 TESLA) 7 Tesla (CMRR) Snyder, C. et al MRM 61(3): 517 -524 (2009)

7 TESLA: A 16 -channel combined loop-dipole transceiver array for 7 Tesla body MRI

7 TESLA: A 16 -channel combined loop-dipole transceiver array for 7 Tesla body MRI Ertürk et al. MRM (2016). 26

Anatomical contrast @ 7 T T 1 W T 2 W SWI STN SN

Anatomical contrast @ 7 T T 1 W T 2 W SWI STN SN RN STN: subthalamic nucleus SN: substantia nigra RN: red nucleus SWI: Susceptibility-weighted imaging Abosch, Yacoub, Ugurbil, Harel. 2010

Susceptibility-Weighted Imaging @ 7 T STN SN Magnet: 7 T Resolution: 0. 4 x

Susceptibility-Weighted Imaging @ 7 T STN SN Magnet: 7 T Resolution: 0. 4 x 0. 8 mm STN = Subthalamic Nucleus SN = Substantia Nigra Dr. Noam Harel, University of Minnesota / CMRR

Susceptibility-Weighted Imaging @ 7 T i P G GP = Globus pallidus e P

Susceptibility-Weighted Imaging @ 7 T i P G GP = Globus pallidus e P G Pi e P G SWI @ 7 T (In-vivo) G Schaltenbrand Wahren Atlas (Ex-vivo) Lamina pallidi mediali s Abosch, Yacoub, Ugurbil, Harel. 2010

DBS STN SN Red. N Dr. Noam Harel, University of Minnesota / CMRR

DBS STN SN Red. N Dr. Noam Harel, University of Minnesota / CMRR

Orientation ODC Phase Left Right 1 mm Yacoub, Shmuel, et al. Yacoub, Harel, Uğurbil

Orientation ODC Phase Left Right 1 mm Yacoub, Shmuel, et al. Yacoub, Harel, Uğurbil Neuroimage (2007) 37(4): 1161 -77 PNAS (2008) 105(30): 10607 -12

Tonotopic Mapping in Human Primary Auditory Cortex 7 T GE f. MRI 1. 2

Tonotopic Mapping in Human Primary Auditory Cortex 7 T GE f. MRI 1. 2 x 1. 5 x 2. 4 mm 3 CMRR/U Maastrich

Cortex has known neural networks that span cortical layers

Cortex has known neural networks that span cortical layers

High Resolution Frequency maps (tonotopy) in the Human Brain at 7 Tesla Primary Auditory

High Resolution Frequency maps (tonotopy) in the Human Brain at 7 Tesla Primary Auditory Cortex Laminar Resolution De Martino et al (PNAS 2015) Formisano et al. (Neuron, 2003) Medial Geniculate Body Moerel et al (Scien. Rep. , 5; 17048 (2015)) Inferior Colliculus De Martino et al. (Nature Communications, 2013)

MT Flattened maps of 3 layers surface Zimmermann et al. PLo. S ONE 6(12):

MT Flattened maps of 3 layers surface Zimmermann et al. PLo. S ONE 6(12): e 28716. (2011) deep CMRR & Univ. Maastricht Collaboration 35 PLo. S Zimmermann et al. 0. 8 mm isotropic 6(12): e 28716. (2011)

7 TESLA HUMAN BRAIN 1 H NMR spectrum at NAA STEAM TE = 6

7 TESLA HUMAN BRAIN 1 H NMR spectrum at NAA STEAM TE = 6 ms TR = 5 s VOI = 8 ml NT = 160 Cr PCr residual water Asp GABA Glc Cr PCr scyllo-Ins Glu Gln Cho GSH Glu Tau Ins Gln NAA Ins 36

lactate: resonates at 1. 3 ppm lipids: resonates at 1. 3 ppm alanine: resonates

lactate: resonates at 1. 3 ppm lipids: resonates at 1. 3 ppm alanine: resonates at 1. 48 ppm N-acetylaspartate (NAA): resonates at 2. 0 glutamine/glutamate: resonates at 2. 2 -2. 4 ppm GABA: resonates at 2. 2 -2. 4 ppm 2 -hydroxyglutarate: resonates at 2. 25 ppm 6 citrate: resonates 2. 6 ppm creatine: resonates at 3. 0 ppm choline: resonates at 3. 2 ppm myo-inositol: resonates at 3. 5 ppm water resonates at 4. 4 ppm 37

7 T, STEAM TR/TE=5000/6 ms VOI=20 x 22 x 20 mm 3 32 scans

7 T, STEAM TR/TE=5000/6 ms VOI=20 x 22 x 20 mm 3 32 scans During visual stimuli, changes of metabolite concentrations were within ± 0. 2 mmol/g 38

Results - 5. 3 min stimulation Group Analysis Error bars = CRLB 39

Results - 5. 3 min stimulation Group Analysis Error bars = CRLB 39

SPINOCEREBELLAR ATAXIAS N-acetylaspartate (NAA): resonates at 2. 0 40

SPINOCEREBELLAR ATAXIAS N-acetylaspartate (NAA): resonates at 2. 0 40

UHF Prostate GOAL: Develop the imaging hardware and methods to exploit the advantages of

UHF Prostate GOAL: Develop the imaging hardware and methods to exploit the advantages of UHF imaging in the prostate. • Development of novel RF coils – Surface arrays – Multi-channel endorectal Coils • Improved Imaging and Spectroscopy – Increased resolution anatomic imaging – Increased resolution functional imaging – Improved spectral quantification on spectroscopy • Future Work: evaluate the potential advantages in cancer detection and staging to improve the identification of clinically significant disease.

3 T 7 T Sagittal T 2 Coronal T 2 Axial ADC

3 T 7 T Sagittal T 2 Coronal T 2 Axial ADC

7 T Intro Our work on 7 T (UHF) breast MR has focused on

7 T Intro Our work on 7 T (UHF) breast MR has focused on RF coil development. We’ve made 4 iterations of RF coils, and while we’ve used them to get some prelim data they are all unsatisfying. The problem is that for contrast-enhanced breast imaging you need relatively uniform flip angles over both breasts and axillae. With the reduced RF wavelength at 7 T (~12 cm) it is very difficult to achive this. This is why we’ve had more success with smaller regions (prostate, cardiac, kidneys) that are more tolerant to flip angle variation.

7 T Breast MR Imaging High SNR allows high-resolution anatomic imaging Axial fat-sat 3

7 T Breast MR Imaging High SNR allows high-resolution anatomic imaging Axial fat-sat 3 D FLASH 0. 5 x 2 mm Sagittal fat-sat 3 D FLASH 0. 3 x 2. 2 mm

7 T Breast MR Spectroscopy Increased spectral resolution and SNR gives excellent spectra in

7 T Breast MR Spectroscopy Increased spectral resolution and SNR gives excellent spectra in normals and patients Patient with IDC Normal Volunteer 6 4 ppm 2 0 Haddadin I. et al. , NMR Biomed 2009 6 4 ppm 2 0 Bolan P. J. et al. , ISMRM 2006 choline: resonates at 3. 2 ppm

3 T Advancements • The following slides are 3 T only: they describe the

3 T Advancements • The following slides are 3 T only: they describe the development of highresolution diffusion imaging

High-resolution Breast DWI: RS vs SMS-IPSE T 2 -weighted anatomical Comparing approaches with fixed

High-resolution Breast DWI: RS vs SMS-IPSE T 2 -weighted anatomical Comparing approaches with fixed 5 min acquisition time 0. 8 x 3 mm Standard b = 0 1. 7 x 4 mm RS-EPI b = 0 1. 8 x 2. 4 mm SMS/MB-IPSE b = 0 1. 25 x 2. 5 mm • Std, single-shot EPI • RESOLVE (RS-EPI), following Wisner JMRI 2014 • SMS/MB with in-plane slice encoding SMS/MB-IPSE approaches T 2 resolution & quality

Renal Blood Flow (m. L/100 g/min) 400 0 Li, NMR in Biomed: 63 -69;

Renal Blood Flow (m. L/100 g/min) 400 0 Li, NMR in Biomed: 63 -69; 2015 Li (2015) ISMRM #3155.

NIH invests substantially (~7 -8 billions/year) in Neurosciences and Brain disorders Research NIH Brain

NIH invests substantially (~7 -8 billions/year) in Neurosciences and Brain disorders Research NIH Brain Research through Advancing Innovative Neurotechnologies (BRAIN) Initiative 49

Description of the functional and structural connections among gray matter locations in the human

Description of the functional and structural connections among gray matter locations in the human brain • spontaneous fluctuations in an f. MRI time series (i. e. “Resting State” f. MRI ) to deduce ‘functional connectivity’ and/or • Diffusion weighted MRI to infer ‘structural connectivity’ 50

Principal Investigators: David C. Van Essen and Kamil Ugurbil 51

Principal Investigators: David C. Van Essen and Kamil Ugurbil 51

“FUNCTIONAL CONNECTIVITY in the HUMAN BRAIN THROUGH • spontaneous correlated fluctuations in an f.

“FUNCTIONAL CONNECTIVITY in the HUMAN BRAIN THROUGH • spontaneous correlated fluctuations in an f. MRI time series (i. e. “Resting State” f. MRI ) to deduce ‘functional connectivity’ Time Spatial patterns of correlated temporal dynamics, resembling 52 activation maps

FUNCTIONAL CONNECTIVITY, DENSE CONNECTOME (from Multiband Resting State f. MRI, 3 T Skyra Connectom)

FUNCTIONAL CONNECTIVITY, DENSE CONNECTOME (from Multiband Resting State f. MRI, 3 T Skyra Connectom) Functional connectivity map (location 1) Functional connectivity map (location 2) Correlation Low High M. Glasser , D. Van Essen et al for the HCP

CMRR 7 T DWI UCLA tractography 54

CMRR 7 T DWI UCLA tractography 54

CMRR 7 T DWI UCLA tractography 55

CMRR 7 T DWI UCLA tractography 55

Wash. U-UMinn-Oxford Consortium. Insan Konnektom Projesi

Wash. U-UMinn-Oxford Consortium. Insan Konnektom Projesi

Nat Neurosci 2015; 18(11): 1565 -1567. 57

Nat Neurosci 2015; 18(11): 1565 -1567. 57

Correlation Between each Subject Measures (SM) and Canonical Correlation Analysis (CCA) Mode 58

Correlation Between each Subject Measures (SM) and Canonical Correlation Analysis (CCA) Mode 58

Correlation Between Subject Measures (SM) and Canonical Correlation Analysis (CCA) Mode 59

Correlation Between Subject Measures (SM) and Canonical Correlation Analysis (CCA) Mode 59

60

60

LIFESPAN - HUMAN CONNECTOME PROJECT • BABY CONNECTOME (birth to 5 years) • DEVELOPING

LIFESPAN - HUMAN CONNECTOME PROJECT • BABY CONNECTOME (birth to 5 years) • DEVELOPING CONNECTOME ( Ages 5 -21) • AGING (Ages 36 - 90+)

SCALES OF BRAINBOW approach to uniquely label individual cells Human Brain Functional Connectivity 62

SCALES OF BRAINBOW approach to uniquely label individual cells Human Brain Functional Connectivity 62

10. 5 T/88 cm; 110 Tons Stored Energy 280 MJ 10. 5 T Conductor

10. 5 T/88 cm; 110 Tons Stored Energy 280 MJ 10. 5 T Conductor length 1160 kms http: //www. healthtalk. umn. edu/2013/12/31/move-110 -ton-imaging-magnet/

68

68

10. 5 Tesla Images of Porcine head Geometrically Adjustable Dipole Coronal TSEs tse_cor_te=67 ms_bw=500_512_Vref=78

10. 5 Tesla Images of Porcine head Geometrically Adjustable Dipole Coronal TSEs tse_cor_te=67 ms_bw=500_512_Vref=78 seq = tse 2 d 1_10 SE SKSPOSP Opt: TR/TE = 5000. 00 / 71. 00 ms nominal flipangle = 130. 00 deg Variable: N Excitation Voltage = 48. 05 V Ref. Voltage: 78. 00 V thk = 3. 00 mm etl = 10 ; TA 04: 16 512*512; Fo. V 170*170; Resolution = 0. 33 x 0. 33 mm G. Adriany, P-F. , Van de Moortele et al.

Towards imaging the body at 10. 5 Tesla Swine post-mortem Phase Shim only Swineimaging,

Towards imaging the body at 10. 5 Tesla Swine post-mortem Phase Shim only Swineimaging, (75 kg), post-mortem; A. Ertürk, Y. Eryaman, G. Metzger

Towards imaging the body at 10. 5 Tesla Swine (75 kg), post-mortem; Phase Shim

Towards imaging the body at 10. 5 Tesla Swine (75 kg), post-mortem; Phase Shim only A. Ertürk, Y. Eryaman, G. Metzger

Log Millimeters 2 1 UHF Functional Imaging with Magnetic Resonance (f. MRI) 0 -1

Log Millimeters 2 1 UHF Functional Imaging with Magnetic Resonance (f. MRI) 0 -1 -2 -3 Calcium or Voltage Imaging with LIGHT

4 Tesla (Installed in CMRR in ~1990)

4 Tesla (Installed in CMRR in ~1990)

Center for Magnetic Resonance Research Univ. of MINNESOTA 74

Center for Magnetic Resonance Research Univ. of MINNESOTA 74

Concept: Future Head MRI PI: M. Garwood, and T. Vaughan (CMRR, Uminn) Small magnet

Concept: Future Head MRI PI: M. Garwood, and T. Vaughan (CMRR, Uminn) Small magnet size dictates • The use of high temperature superconductors • Imaging at inhomogeneous magnetic fields