MR physics 101

MRI Physics 101

Prof. Sairam Geethanath, Ph.D. Medical Imaging Research Centre

Dayananda Sagar Institutions 1st October 2016

A"en

ua'o

n(E/E0)

(µV)

γ2G2δ2(∆-δ/3)(109sm-2)

Declara'on

•  Ihavenoconflictsofinteresttodeclare,withrespecttothecontentsofthispresenta'onanddatashownhereisalreadyinpublicdomain

• WiproGEHealthcarefundspartofmyresearchaspartofaDSTTechnologysystemsdevelopmentgrant

MIRC 2

Outline•  Highschoolphysics–relevanttoMRI

•  OverviewoftheMRIsystem

•  PrecessionandthePeltonWheel

•  Relaxa'on'mes:T1andT2

•  Basisforcontrastgenera'on

•  SpinEchosequence:Anexampleofimagegenera'on•  Spa'allocaliza'onthroughmagne'cfieldgradient•  k-spaceanditstraversal

•  Toolstogetstarted•  Pulseq-Acquisi'on•  GPI-Reconstruc'on

•  DIYassignmentsMIRC 3

ImportanthighschoolPhysics

4

Lenz’slaw Reciprocitytheorem BiotSavart’sLaw Faraday’slaw

Vineet,MIRC Abitha,MIRC Aparana,MIRC Anusha,MIRC

Gradientcoils

Subject

Radiofrequency

coil

Magnet

OverviewofaMRIsystem

Imagecourtesy:MRIscannercutaway:Colinmcnulty.com

BOLD

Imagecrea'on

Birdcagecoil

Aliasing

Baselinecorrec'on

Blochequa'on

CPMGSNR Angiography ADC DWI TE/TR Eddycurrent

ContrastbasisSWI

FID

Gyromagne'cra'o

Fouriertransform

Echoes

Keyholeimaging

DCE-MRI

Larmorfrequency

T1/T2relaxa'on

Representa8onofatypicalMRIscannerMIRC 5

Larmor Equation

EffectoftheB0,G.randB1fields

z

x

y

MIRC 6

Whydoweneedto'pthespins?–Understandingprecession

•  The application of an RF pulse, causes alignment of spins away from the longitudinal axis (lower energy state) on a transverse plane (higher energy state)

•  Spins release the absorbed energy and drop back to their lower energy states •  Spin can exchange a quantum of energy with the lattice (also precessing at same frequency)

•  Spin transitions from 𝑚=− 1/2  (excited state) to 𝑚= 1/2  (ground state) is accompanied by a transition upwards in energy from some lower lattice state to a higher lattice state h

•  Energy transition must be equal ‘ law of conservation of energy’ •  Transfer of energy occurs through collisions, rotations, or electromagnetic interactions with the surrounding lattice •  This energy loss is unrecoverable and represents the transfer of heat.

h"p://mriques'ons.comMagne'cresonanceimaging:Physicalprinciplesandsequencedesign

𝑚=− 1/2 

𝑚= 1/2 

ℎ

𝑙

𝑝𝑟𝑜𝑡𝑜𝑛 𝑙𝑎𝑡𝑡𝑖𝑐𝑒

Boltzmannequa'onforpopula'onstates

T1 relaxation

•  The electromagnetic field from a particle can be considered to emanate from an idealized tiny bar magnet with north and south poles ("dipole") .

•  A dipole-dipole interaction is a "through space" interaction of the fields from two spinning particles

•  Four major factors determine the strength of the dipolar interaction: (1) types of spins; (2) the distance between them; (3) the angle between them; and (4) their relative motion.

h"p://mriques'ons.com/dipole-dipole-interac'ons.html

T1ofwater T1ofwaterdopedwithCopperSulphate T1ofoil

Dipole – dipole interactions

h"p://mriques'ons.comMagne'cresonanceimaging:Physicalprinciplesandsequencedesign

0 2 44.0

4.5

5.0

Mz

t

M0

FirstRF SecondRF

TR

T1

0.63M0

•  Spinlamceinterac'onresultinre-growthoflongitudinalmagne'za'on

•  Rateofchangeoflongitudinalmagne'za'oniscapturedbyanexponen'alrecovery,isacrossproductofmagne'za'onmomentMandtheappliedexternalfieldB

•  Synonyms:longitudinalrelaa'on,thermalrelaxa'onandspin-lamcerelaxa'on

T1 recovery

Spin-SpinandEffec-veSpin-Spinrelaxa-on:•  NMRsignal–phasecoherenceofnuclearspins&Exponen'alsignaldecay–lossofphasecoherence

•  Spin-spinrelaxa'on–dipolecouplingbetweenneighbouringspins

•  Sampledsignal,asafunc'onof'meisgivenby, S(t)=S0exp(-t/T2)Where,S0–ini'alsignalmagnitudeatt=0

•  Generally,thefieldisnoten'relyhomogeneous

•  Phasecoherenceloss–combina'onofspin-spinrelaxa'onandmagne'cfieldinhomogeneity,whichintroducearangeofLarmorfrequencies

•  Dispersionoffrequencies–lossofcoherence-signaldecay

Incompletelyhomogeneousfield,thephasecoherencelossisduetospin-spinrelaxa8on

T2 relaxation

•  Effec'vespin-spinrelaxa'on'meconstant,T2*isdefinedas,1/T2∗ = 1/T2 +γ∆B0

Where,

γ–gyromagne'cra'ooftheobservednucleus

∆B0–magne'cfieldinhomogeneity•  Signalasafunc'onof'meinsitua'onwithfieldinhomogeneityisgivenby,

S(t)=S0exp(-t/T2∗)Where,

S0–ini'alsignalmagnitudeatt=0

Phasecoherencelossduetospin-spinrelaxa8onisirreversible,whereaslossduetofieldinhomogeneitycanbereversedbySpinEcho

(calledasHahnEcho)ErwinHahn

T2* relaxation

Figure1:Spin-Echopulsesequencediagram

•  The'mebetweenthe90°pulseandthe180°pulseiscalledTΕ,theecho-8me

•  Reversingthede-phasingduetomagne'cfieldinhomogeneityisthegoalofthisbasicspin-echoexperiment

•  Onlythede-phasingthatoccurredasaresultofmagne'cfieldinhomogeneitywillbere-focused

spin echo

TerranovastudentGuide,MagritekLimited

•  Samplingcommencesatthecentreoftheecho

•  Delaybetweenthe180°pulseandthefirstsampleddatapointisTΕ

•  TΕmustbechosentobelongenough -toviewtheen'reecho -toallowforthecompleterelaxa'onofthesignalexcited

T2Measurement:•  Measuredusingasuccessionofspin-echoexperimentswithincrementallylongerecho'mes

•  Theplotofechoamplitudeasafunc'onofecho'mewillbeanexponen'aldecaywithacharacteris'cdecay'meconstant,T2

•  Theechoamplitudeisgivenby,

whereE-amplitudeofanechoacquiredwithTΕ E0-echoamplitudeintheabsenceofaT2decay

Relaxa'onprocesses–Blochequa'ons&contrast

MIRC 15

h"ps://www.chemie.uni-hamburg.de/nmr/insensi've/tutorial/en.lproj/vector_model.htmlh"p://www.dayanandasagar.edu/index.php/sharing

Bloch,1946

T1 T2 DIFFUSION

Spa'allocaliza'on–Lauterburgradientexperiment

MIRC 16Lauterbur,Nature,1973

Theory

ThesignalacquiredistheFreeInduc'onDecay(FID)fromallthespinsofthepar'cularslice

TheLarmorfrequencyisgivenby

f

t

A

A

DiscreteFourierTransform

Avisualrepresenta'onofk-space

Topviewofk-space

• Idealk-spaceisHermi'aninnature,discoun'ngerrorsfrommeasurement

• Usedinacquisi'onslikeHASTE

JargonforengineersJ

Frontview Topview

Ananalogyfork-spacetrajectory

Considerahill

20 40 60 80 100 120

50

100

150

200

250

Fewk-spacetrajectoriestrajectories

y

x

Exampleproblem1:DesignCartesiank-spacetrajectory

Givenparameters: ∆x = 1 mm ∆y = 1 mm Lx = 25.6 cm L y = 25.6 cm

Evaluatetheunknowns:

Variable Value

Nx

Ny

256

256 3.9 m-1

3.9 m-1

[-500, 500] m-1

[-500, 500] m-1

kx

ky

RF Pulse

Gz

Gy

Gx

TimingDiagramDepic'ngCartesianSampling

Gx

Gy

Gz

RF Pulse

Reconstruc'on&ar'facts

OpensourcetoolforPSD-Pulseq

MIRC 27

0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.090

500

1000

1500

RF

mag

(Hz)

0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09t (s)

0

2

4

RF

ph (r

ad)

0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09t (s)

-500

0

500

Gz

(kH

z/m

)

0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09-600

-400

-200

0

Gy

(kH

z/m

)

0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09-1000

0

1000

Gx

(kH

z/m

)

0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09-1

0

1

ADC

Source:h"p://pulseq.github.io/(NeedsMatlabinstalled)Laytonet.al.,MRM2016

Opensourcetoolforstudentstouse–recon-GPI

MIRC 28

•  h"p://gpilab.com/2016/09/gpi-on-windows/

•  Simula'on

•  Reconstruc'on

•  ImageProcessing

•  Visualiza'on

•  MacOSX/Ubuntu/Windows

Zwartet.al.,MRM,2016

DIYassignments

MIRC 29

Courseassignments(andsolu'ons):1. Introduc'onandpreliminaries2. Physics3. WORKINPROGRESS(IntroductoryPSDandrecon)4. FastImaging5. MRapplica'ons

? MIRC 30

[email protected] http://dayanandasagar.edu/index.php/mirc http://dayanandasagar.edu/index.php/sharing

Prof. Sairam Geethanath, Ph.D. Medical Imaging Research Centre Dayananda Sagar Institutions 1st October 2016

¡  Head¡ Tumors,aneurysms,bleedinginthebrain,¡ Nerveinjury,damagecausedbystroke.

¡  Spine¡  Discs and nerves of the spine for conditions such

asspinalstenosis,discbulges,andspinaltumors.

¡  Chest:Heart,thevalves,andcoronarybloodvessels

¡  Bloodvesselsand?low–Dr.RameshVenkatesan’stalk

¡  Abdomenandpelvis¡  Belly,liver,gallbladder,pancreas,kidneys,andbladder

¡  Bonesandjoints¡ Arthritis,problemswithjoints,bonemarrowproblems,¡ Bonetumors,cartilageproblems,¡ Tornligamentsortendons,infection.

¡  DWIisacommonlyusedMRIsequenceforevaluationofacuteischaemicstroke

¡  Sensitiveindetectionofsmallandearlyinfarcts

¡  Noninvasivewayofunderstandingbrainstructuralconnectivityandmacroscopicaxonalorganization

¡  DWimageisgeneratedbasedonthedirectionalrateofdiffusionofwatermoleculesinsidethebrainduetoBrownianmotion

¡  Imageintensitiesinverselyrelatedtotherelativemobilityofwatermoleculesintissueandthedirectionofthemotion

Result Interpretation

Quantitative analysis

Segmentation

Feature Extraction

Visualization (may include generating FA,ADC, Tractograpghy)

Processing (may include DWI enhancement using Super resolution techniques )

Preprocessing (may include registration,skull stripping,normalization motion , denoisng of low field MR

image/DWI)

¡  Dark regions – water diffusing slower, more obstacles to movement OR increased viscosity

¡  Bright regions – water diffusing faster DWI

¡  Bright regions – decreased water diffusion

¡  Dark regions – increased water diffusion

Figure Source: www.radiopaedia.org

ADC

Matlab Code available with Arush, MIRC

COLOUR FA MAP

TRACTOGRAPHY

•  According to the principal direction of diffusion, colour coding of the diffusion data is done

•  Red - transverse axis (x-axis) •  Blue – superior-inferior (z –axis) •  Green – anterior-posterior axis (y-

axis)

•  I n t en s i t y o f t he c o l ou r i s proportional to the fractional anisotropy

•  It is 3D modeling technique used to visually represent neural tracts using data collected by diffusion tensor imaging (DTI)

•  Voxels are connected based upon

similarit ies in the maximum diffusion direction.

Figure Source: www.radiopaedia.org

Matlab Code available with Arush, MIRC

ADC map computa'on

b=0 b=100 b=200 b=500 b=1000

Signalintensitydecreasingwithincreaseinb-value ADCmapscanner ADCmapmatlab

FAmap MeanDiffusivitymap

Scanner

MATLAB

ColorcodedFAmap

Frac-onalAnisotropymap

FA ColFA

Nofilter

PMfilterusingfixedKappa

PMfilterwithAutomatedKappa

Scannerresults

DiffusionWeightedimagesdenoisingforFAmapcomputa-on

Figure source: Nucifora et al. Radiology 245:2 (2007)

Corticospinal Tracts -Probabilistic Corticospinal Tracts - Streamline

1.Streamline tractography •  Connects neighbouring voxels

from user defined voxels (seed regions)

•  Tracts are traced until termination criteria are met.

2.Probabilistic tractography •  Value of each voxel in the map is the

probability the voxel which is in the diffusion path between the ROIs.

•  It provides quantitative probability of connection at each voxel

•  Allows tracking into regions where there is low anisotropy.

Degree of anisotropy Streamline tractography Probabilistic tractography

Figure source :Nucifora et al. Radiology 245:2 (2007)

•  Ithasbeenwellestablishedthatmagne'cresonanceimaging (MRI) provides cri'cal informa'on aboutcancer[3]

•  Magne'c resonance spectroscopic imaging (MRSI)furthers this capability by providing informa'onabout the presence of certain ‘metabolites’ whichare known to be important prognos'c markers ofcancer [4] (stroke, AD, energy metabolism, TCAcycle)

•  MRSI provides informa'on about the spa'aldistribu'on of these metabolites, hence enablingmetabolicimaging

[3]HukWJetal.,NeurosurgicalReview7(4)1984;[4]PreulMCetal.,Nat.Med.2(3)1996;

Metabolicimaging:applica'ons

CANCER

NORMAL

[5]HKugeletal.,Radiology183June1992 MIRC 43

[5]

•  3D- PRESS makes it possible to localize the signal in the voxel formed by the intersection of the three slices

Figure7:Displayofthevolumeofinterest(voxel)locatedattheintersec8onoftheslices

[3]*

SpectroscopicImagingMethods•  SpectroscopicImagingmethodsmapspa'aldistribu'onofcomponentswithdifferentchemicalshi}s•  Thisimagingisacombina'onofspa'aladspectralimaging•  Goal:ToobtainNMRspectrumateachspa'alposi'on/todisplayanimageofeachchemicalshi}1.3DFourierTransformSpectroscopicImaging:

RF

GZ

Gy

Gx

DAQ

Figure4:3DFourierTransformSpectroscopicImagingSequence,Phaseencodinginxandyisfollowedbydataacquisi8on(DAQ)withgradients

turnedoff

1*

RF

GZ

Gy

Gx

DAQ

TE

90 180

Figure5:3DFTSpinEchoSpectroscopicImagingSequence

2.SpectroscopicImagingwithTime–VaryingGradients•  Signalisreadoutinpresenceofgradientrota'ngatangularfrequencyΩasinfigure6

•  Usingthesegradients,datacanbeacquiredoverarangeoffrequenciesthusavoidingaliasing

Figure6:SpectroscopicImagingSequencewithRota8ngGradients

tRF

GZ

Gy

Gx

DAQ

sin Ωt

cos Ωt

1*

•  Longacquisi'on'mesforMRSI•  AtypicalMRSIprotocol(32X32X512)takes~10-12minutes•  Difficulttomaintainanatomicalpostureforlong'me•  Increases pa'ent discomfort, likelihood of early termina'on ofstudy

•  Discouragesrou'neclinicaluseofthispowerfulMRItechnique

•  Toincreasethroughput(decreasedscanner'me,technician'me)

•  Reduc'on of acquisi'on 'me is usually accomplished byunder-samplingmeasureddata(k-space)

•  Limita'onsofShannon-Nyquistcriterion

•  Compressed sensingprovidesa framework toachieve sub-Nyquistsamplingrateswithgooddatafidelity

CS-MRSI:Needforaccelera'on

MIRC 47

kx

ky

x

y

3DFT

Brain - normal (N=6)

Brain - cancer (N=2)

Prostate -cancer (N=2)

MRSI data Scanner TR(ms) TE(ms) # Averages Grid Size FOV (mm3)

Brain - normal (N=6)

Siemens 3.0T Trio Tim 1700 270 4 16 x 16 x 1024 100 x 100 x 15

Brain cancer (N=2) Philips 3.0T Achieva 1000 112

112 2 2

18 x 21 x 1024 19 x 22 x 1024

180 x 210 x15 190 x 220 x 15

Prostate cancer (N=2) Philips 3.0T Achieva 1200

1000 140 140

1 1

14 x 10 x 1024 16 x 12 x 1024

25 x 50 x 33 20 x 51 x 26

Insilicoandinvitrophantomstudiesreportedin[6]Geethanathet.al.,SPIEMedicalImaging2010[7]Geethanathetal.,Radiology.2012

MRSI:acquisi'onparameters

MIRC 48

[7]

Applica'onofCStoMRSI

MIRC 49

•  Signalmodelofafree-induc'ondecaywithN(3inthiscase)metabolites

•  ThesparselymeasuredFourierdataisrepresentedbyy,Objecttobees'matedisin(x,y,f)spaceism

•  Undersamplinginx-ydimensionsvsx-fdimension•  Problemdefini'on:

•  FindthesparsesttransformcoefficientsofmthatprovidesfordataconsistencybetweenFouriercoefficientsmeasuredandes'mated,atsampledloca'ons

(2) argminm∥Fu(m)-y∥2

2+λ∥ψ(m)∥1 (3)

(1)

Processing So=wares: [1] jMRUI:

It is a software that can be used to process MRSI data.

The spectra are typically subjected to the following processing steps in jMRUI [5]:

(a) Apodization to remove existing truncation artifacts,

(b) baseline correction,

(c) time-domain Hankel-Lanczos singular value decomposition filtering of residual water and fat peaks,

(d) Phase Correction,

(e) Frequency Shift.

5*

[2] VeSPA:

It is a open source software for MRS applications. It supports four applications:

1. RFPulse (for RF pulse design),

2. Simulation (for spectral simulation),

3. Priorset (for creating simulated MR spectroscopic data)

4. Analysis (for spectral data processing and analysis)

6*

Cr23.916

Cho3.186

Cr3.03

NAA2.008

Lipids0.9-1.4

Gln,Glu,GABA2.12-2.42

Figure8:SpectrasimulatedusingVeSPAsoVwarewithmajormetabolitesofbrain

[7]Geethanathetal.,Radiology.2012

MRSI:invivonormalbrain

1X

NAA Cr Cho

5X

MIRC 52

[7]Geethanathetal.,Radiology.2012

Brain cancer

1X

2X

5X

10X

Prostate cancer Normal Cancer Normal Cancer

NAA Cr Cho NAA

Cr Cho Cr2 Cr2 Cr

Cho Cit Cit

Cho + Cr

CS-MRSI:Cancerresults

MIRC 53

Brain - cancer

Prostate - cancer

Brain - Normal

Brain - Normal Brain - cancer

Prostate - cancer

CS-MRSI:Metabolitemaps

MIRC 54[7]Geethanathetal.,Radiology.2012

Limita'ons of PRESS •  Conventional slice-selective 180 refocusing pulses do not have particularly good slice profiles, leading to

non-uniform metabolite excitation and signal generation from outside PRESS box

•  By definition, it restricts excitation to a rectangular volume, but brain has a curved, elliptical shape – difficult to obtain signal from cortical regions close to the skull

•  3DPRESS-MRSI sequence - scan-time becomes very long if high spatial resolution in all three directions is required (number of PE gradients to be recorded becomes very high since encoding is performed in all three directions, hence giving long scan times)

•  Difficulty of obtaining sufficient magnetic field homogeneity for large spatial coverages

[4]*

Schizophrenia: - In a study [7], it is shown that using proton MRSI, in case of patients with schizophrenia, there will be a relative loss of signal from N-acetyl- containing compounds (NAA)

- Patients with schizophrenia, when compared as a group to normal controls, show a consistent 1H-MRSI pattern of group differences, i.e., bilateral reductions of NAA/CRE and NAA/CHO in HIPPO and DLPFC;

- 1H-MRSI data in both patients and controls do not show significant changes over a period of 90 days; however, absolute metabolite ratios in individuals show low predictability over this time interval;

- 1H-MRSI data show relatively low variability (as measured by the coefficients of variation (CVs)) both in patients and normal controls, especially for NAA/ CRE and CHO/CRE.

7*

Mild Cogni've Impairment: - Mild cognitive impairment (MCI) is a clinical state between normal aging and Alzheimer's disease (AD)

- In a study [8], 1H-MRS findings were compared in the superior temporal lobe, posterior cingulate gyri and medial occipital lobe among 21 patients with MCI, 21 patients with probable AD, and 63 elderly controls

- Results showed that, NAA /Cr ratios were significantly lower in AD patients compared to both MCI and normal control subjects in the left superior temporal and the posterior cingulate VOI

- Myoinositol (MI) /Cr ratios measured from the posterior cingulate VOI were significantly higher in both MCI and AD patients than controls

- Cho /Cr ratios measured from the posterior cingulate VOI were higher in AD patients compared to both MCI and control subjects

8*

•  Increasedinforma'oncontentbutatcostofincreasedacquisi'on'me

•  Provides richer insight into the pathophysiology and direct impact on therapeu'cdesign

•  Mul'pleopen-sourcetoolsavailable–jMRUIandVespa

•  Increasedclinicalresearchinneuro-,breast,prostate,cardiac(murine)andliver

•  Ac'veareaofresearch–developmentofPSDandrecon

Summary

MIRC 58

References [1] Nishimura, Dwight George. Principles of magnetic resonance imaging. Stanford University, 1996.

[2] SG Dissertation

[3] Theoretical background: MRI and MRS

[4] Peter B. Barker et al., “In vivo proton MR spectroscopy of the human brain”, Progress in Nuclear Magnetic Resonance Spectroscopy 49 (2006) 99–128

[5] A. Naressi, et al. Java-based graphical user interface for MRUI, a software package for quantitation of in vivo/medical magnetic resonance spectroscopy signals. Computers in Biology and Medicine, 31(4), 269-286 (2001)

[6] https://scion.duhs.duke.edu/vespa/

[7] Alessandro Bertolino et al., “Reproducibility of Proton Magnetic Resonance Spectroscopic Imaging in Patients with Schizophrenia”, Neuropsychopharmacology 1998

[8] K. Kantarci et al., “Regional Metabolic Patterns In Mild Cognitive Impairment And Alzheimer's Disease A 1h Mrs Study”, Neurology. 2000

Acknowledgement• People

§ Prof.VikramD.Kodibagkar§ Prof.JosephV.Hajnal§ ColleaguesatUTSouthwestern§ ColleaguesatICL§  StudentsatMIRC&radiologistsatSagarHospital

§ Collaboratorsfrom§  UMN§  Oxford§  GEHealthcare(Dr.RameshVenkatesan)§  ASU§  IISc§  NIMHANS§  PhilipsHealthcare

•  Funding

§ Pilotgrant(PI:Kodibagkar)fromUL1RR024982,(PI:MiltonPacker)

§ ARP#010019-0056-2007(PI:Kodibagkar)§ R21CA132096-01A1(PI:Kodibagkar)§ W81XWH-05-1-0223(PI:Kodibagkar)§ R21CA139688(PI:Corum)§ S10RR023730(PI:Garwood)§ P41RR008079(PI:Garwood)§ MRIIndiaNa'onalMissiongrant–SCANERA(co-PI:Geethanath)fromDEITY§ DST-TSDgrantandWiproGEHealthcare(PI:Geethanath)§ KFISTgrant(PI:Geethanath)

MIRC 60

? MIRC 61

[email protected] http://dayanandasagar.edu/index.php/mirc http://dayanandasagar.edu/index.php/sharing

References

•  [1]M.Kassetal.Interna'onalJournalofComputerVision,1988.•  [2]HandbookofMRIPulseSequences,M.A.Bernstein•  [3]M.GrantandS.Boyddisciplinedconvexprogramming.•  [4]M.Lus'nget.al.MRM,2007•  [5]A.S.Konaretal.,JournalofIndianIns'tuteofScience,2014.

MIRC 62

Acknowledgement• People

§  Dr.VikramD.Kodibagkar§  Dr.JosephV.Hajnal§  StudentsatMIRC

§  MRSIproject§  HyeonmanBaek,Ph.D.

§  Ma"hewLewis,Ph.D.§  SandeepK.Ganji,B.

Tech.§  YaoDing,M.S.§  RobertD.Sims,M.D.

§  ChanghoChoi,Ph.D.§  ElizabethMaher,M.D.,

Ph.D.

•  Funding

§ Pilotgrant(PI:Kodibagkar)fromUL1RR024982,(PI:MiltonPacker)

§ ARP#010019-0056-2007(PI:Kodibagkar)§ R21CA132096-01A1(PI:Kodibagkar)§ W81XWH-05-1-0223(PI:Kodibagkar)§ R21CA139688(PI:Corum)§ S10RR023730(PI:Garwood)§ P41RR008079(PI:Garwood)§ MRIIndiaNa'onalMissiongrant–SCANERA(co-PI:Geethanath)§ DST-TSDgrantandWiproGEHealthcare(PI:Geethanath)§ KFISTgrant(PI:Geethanath)

§  ROICSproject

§  AmareshKonar,M.Tech

§  Shashikala,M.Tech§  Shivaraj,M.Tech§  RashmiRao,B.E.§  BarjorGimi,Ph.D.§  SteenMoeller,Ph.D.§  JuliannaCzum,MD

§  RUSL/Go-Ac'veproject

§  AmareshKonar,M.Tech

§  PavanPoojar,M.Tech§  NutanDev,B.E.§  SmeraLingesh,M.Tech§  RameshVenkatesan,

D.Sc§  Smt.Prema,B.A.§  ShilpaD.,M.S.

MIRC 63

MIRC 64