Magnetic Resonance Spectroscopy: Basic Principles and Selected Applications

Transcription

1 Magnetic Resonance Spectroscopy: Basic Principles and Selected Applications Sridar Narayanan, PhD Magnetic Resonance Spectroscopy Unit McConnell Brain Imaging Centre Dept. of Neurology and Neurosurgery Montreal Neurological Institute McGill University Montreal, Canada

2 Magnetic Resonance Spectroscopy Noninvasively measure the amount of specific chemical compounds (metabolites) present in tissue Can be used to evaluate chemical pathology and/or microscopic damage in disease states Complementary to MRI

3 NMR Basics Nuclei with odd number of protons and/or neutrons nuclear spin angular momentum ( spin ) nuclear magnetic moment Biological tissue rel. sens. hydrogen ( 1 H) 100% phosphorus ( 31 P) 6.6% carbon ( 13 C) 1.6%

4 NMR Basics B 0 M no external field external field B 0 (spin 1/2) B. Pike, MNI

5 Net Magnetization (M) Define coordinate system with z-axis in the direction of the static magnetic field B 0 z M B 0 x y

6 Resonance: Larmor frequency spins precess and exhibit resonance at the Larmor frequency ω = γb where γ = gyromagnetic ratio γ/2π = MHz/T for hydrogen

7 Free Induction Decay 1 z FID Signal Time x FFT y ω 0 stationary (lab) frame of reference Signal spectrum 10 B. Pike, MNI 0 Frequency

8 Resolving multiple signals with different frequencies Fourier Transform Time-domain signal with 3 component frequencies Spectrum (frequency domain)

9 1 H MRS Basic Principles 1 H nuclei resonate at a characteristic frequency dependent on the magnetic field strength B Within a given applied field B, 1 H nuclei in different chemical environments experience a slightly different effective field due to chemical shielding from surrounding electrons

10 Chemical Shielding σ = diamagnetic screening constant B = B o (1-σ) Joseph P. Hornak

11 Chemical Shielding Magnitude of σ depends on local electron density => chemical environment Adjacent atoms Bonds This causes the 1 H nuclei to resonate at slightly different frequencies => chemical shift ν = (γ/2π) B o (1-σ) Joseph P. Hornak

12 Chemical shift: Methanol CH 3 H OH H O C H H δ ppm Chemical shift n n Horizontal axis n Frequency (ppm) Vertical axis n Area proportional to concentration After Joseph P. Hornak

13 Hz vs. ppm The chemical shift when expressed in Hz is B 0 dependent Express relative to the resonant frequency of a standard reference, v ref, typically tetra-methyl silane (TMS) δ ppm = (ν-ν ref )/ν ref x 10 6 Expressed in parts-per-million (ppm) the frequency shift is independent of B 0

14 What is a spectrum? An NMR spectrum is a plot of signal intensity versus chemical shift Parts per million (ppm) Field independent Resonance intensity Chemical shift

15 Spatial localization: Single Voxel PRESS Point RESolved Spectroscopy Slice selective in 3 planes Acquire 2 nd echo TE down to about 30ms on modern MRIs

16 Volume selection: PRESS

17 Volume selection: PRESS

18 Volume selection: PRESS

19 Spatial localization: Single Voxel STEAM STimulated Echo Acquisition Mode Slice selective in 3 planes Trans=>long=>trans Shorter echo times than PRESS (~20ms) Sharper localization Half the signal

20 B 0 homogeneity: Shimming

21 B 0 homogeneity: Shimming

22 B 0 homogeneity: Shimming

23 MR Imaging H 2 O: 70 M 1H-MR Spectroscopy Metabolites: ~ 7 mm

24 MR Imaging H 2 O: 70 M 1H-MR Spectroscopy Metabolites: ~ 7 mm NA Cho Cr ppm

25 Water suppression CHESS CHEmically-Selective Saturation 90 pulse frequencyselective for water, followed by spoiler gradient Destroy the net magnetization of water protons

26 Example (Siemens Sonata) PRESS, TR = 1500, TE = 135 Eddy current compensation Filtering Zero filling Fourier transformation Frequency shift correction Phase correction Baseline correction Curve fitting

27 MRSI: Spectroscopic Imaging Size: 1-2 cc Scan times: min Quantitation: harder MRSI 2D (single or multislice) 3D EPI

28 Long vs. short TE TE 135 TE 30 Fewer metabolites, simpler baseline More metabolites, more complicated baseline

29 Research and Clinical Applications

30 Observable 1 H Metabolites

31 MR Spectroscopy: monitoring neuronal and axonal integrity in vivo NA NAA image Cho Cr From Coyle, JT. (1989) ppm Simmons,M.L (1991) Moffett, J.R. (1991) Bjartmar, (2002)

32 Mechanisms of NA decrease Decreased density of NA axonal loss axonal shrinkage edema Axonal metabolic dysfunction (Demougeot, 2001; Dautry, 2000; Watson, 1998; Kalra, 1998; Cendes, 1997; Vermathen, 1996; Hugg, 1996; De Stefano, 1995)

33 Reversible NAA decreases with mitochondrial dysfunction NAA n=15 n=9 n=6 NeuN Early N-Acetylaspartate Depletion Is a Marker of Neuronal Dysfunction in Rats and Primates Chronically Treated with the mitochondrial Toxin 3-Nitropropionic Acid. Dautry C., et al, J Cereb Blood Flow Metab 2000;20:789-99

34 CHOLINE overview u Cho, PCho and GPCho u precursors of membrane components u Cho and GPCho u products of membrane degradation Increased Cho membrane turnover cellular density Inflammation

35 MRS: Creatine Creatine Phosphocreatine + Creatine Internal standard (if unaffected) Cho Cr ppm

36 MRS: Lactate Lactate Anaerobic metabolism ischemia acute inflammation tumour LA

37 Multiple Sclerosis spectra from inside the lesion ( ) spectra from homologous contr. voxels ( ) Cho Cr NAA La N. De Stefano

38 Chronic MS: Axonal damage in plaques and NAWM Normal Control" MS Patient" NA" NAWM" Lesion" Cho" Cr"

39 Cerebral NA/Cr vs EDSS 88 MS Patients (55 RR, 33 SP) NA/Cr 4 3 Whole MS group RR MS NA/Cr SP MS SROC = p< De Stefano, et al SROC = p< EDSS NA/Cr SROC = p< EDSS

40 Early damage: EDSS < NA/Cr MTr NBV Z scores NC MS NC MS NC MS p< p< p=0.07 (N=57) De Stefano, et al. 2002

41 Monitoring therapy with MRS

42 MS: NA/Cr with IFNB Normal range" Treated" 3.1 NA/Cr Treated" Untreated" * * Untreated" Narayanan et al Time (months) Time (months)"

43 MRS in MS Axonal injury Extends beyond lesions into NAWM (Narayanan 1997, Fu 1998) Begins in the early stages of MS (De Stefano 2001) NA/Cr Accumulates more rapidly in earlier clinical stages of MS than later correlates more strongly with disability in patients with early stage disease than in patients with more severe MS (De Stefano 2001) Can be used to monitor treatment response (Narayanan 2001)

44 Recurrent Oligodendroglioma! Cho image! 2 NA! Cr! R Cho! 3 1 and 3! 2 1 LA! M. Preul

45 TLE: Regions of Interest Intermediate Medial Temporal Lobe (Mid): head and body of hippocampus amygdala hippocampus Posterior Medial Temporal Lobe (Post): tail of hippocampus & axonal projections L. M. Li

46 Lateralization of TLE in 100 patients: Concordance with EEG-clinical findings Neuroimaging Abnormal Lateralizing AM+HF volumes: 88% 83% Mid + Post-TL NAA/Cr: 99% 86% MRSI + MRIVol : 100% 90% L. M. Li

47 Conclusion MRS is a valuable tool to probe the chemical composition of the brain Provides metabolic information complementary to the structural information provided by MRI Powerful tool for research Can be diagnostically helpful in specific situations

48 Douglas L. Arnold, MD G. Bruce Pike, PhD Nicola De Stefano, MD, PhD Mark Preul, MD Li Li Min, MD Joseph P. Hornak, PhD Acknowledgments

49 Questions?