← Back to Main Index

UAB MRS Resources

Magnetic Resonance Spectroscopy material gathering point.
Focus: Multi-nuclear (X-nuclei) applications at 3T (Clinical).

📖 View Glossary & Jargon 📚 View Learning Resources

Technical Configuration

System Hardware

Scanner: 3T Siemens Prisma (Multinuclear)

Coils & Ancillary:

Polarean: Xenon Hyperpolarizer + Xe Vest Coil
PulseTeq: Dual-tuned surface coils
- ¹³C / ¹H Surface Coil (cf. PulseTeq Liver Coil)
- ³¹P / ¹H Surface Coil

General Considerations at 3T

Hardware & Instrumentation Requirements for non-proton imaging on clinical scanners.
- Broadband amplifiers (Multi-transmit capable?)
- Double-tuned coils (e.g., 1H/31P, 1H/13C) for anatomical localization + shimming
- B0 Shimming challenges (2nd order shims essential)
Pulse Sequences & Physics Adapting standard sequences for X-nuclei.
- FID vs. Spin Echo (T2* considerations)
- NOE (Nuclear Overhauser Effect) enhancement - proton decoupling
- SAR (Specific Absorption Rate) limits with decoupling
- Chemical Shift Imaging (CSI) / MRSI vs. SVS (Single Voxel)

31P Phosphorus Spectroscopy

Bioenergetics The primary clinical application of 31P MRS.
- PCr (Phosphocreatine) / ATP ratios - Energy reserve status
- Pi (Inorganic Phosphate) - Metabolic byproducts
- Intracellular pH measurement (chemical shift of Pi)
- Mg2+ concentration (chemical shift difference between ATP resonances)
Membrane Metabolism Markers of synthesis and degradation.
- PME (Phosphomonoesters) - Membrane building blocks (e.g., PC, PE)
- PDE (Phosphodiesters) - Membrane breakdown products (e.g., GPC, GPE)
- Relevance in oncology (turnover rates)

13C Carbon Spectroscopy

Metabolic Flux Analysis Tracking dynamic metabolic pathways.
- Natural Abundance (1.1%) challenges - requires long scan times.
- ¹³C-Glucose infusion studies (TCA cycle rate, Glutamate/Glutamine cycling)
- Glycogen metabolism (Muscle/Liver) - See: Liver Glycogen
Hyperpolarized 13C (Pyruvate) Emerging clinical applications (requires polarizer).
- Real-time metabolic imaging (Pyruvate -> Lactate conversion)
- Warburg effect visualization in tumors

129Xe Xenon Spectroscopy

Hyperpolarized Imaging Using laser-polarized gas for enhanced signal.
- Lung ventilation imaging (Gas phase)
- Dissolved Phase Spectroscopy (The "MRS" component)
Dissolved Phase Peaks Chemical shift separation of compartments.
- Gas peak (0 ppm reference)
- Barrier peak (Tissue/Plasma ~197 ppm)
- RBC peak (Red Blood Cells ~217 ppm)
- Gas exchange metrics (RBC/Barrier ratios) - Pulmonary fibrosis, COPD assessment
Diffusion Imaging / ADC Probing alveolar dimension and microstructural integrity.
- Brownian motion of gas atoms restricted by alveoli walls
- Apparent Diffusion Coefficient (ADC) mapping
- Mean acinar airway dimension (Lm) and morphometry
- Biomarker for Emphysema / COPD and aging ("senile emphysema")

Clinical Protocols & Sequences (Siemens 3T)

Localization Strategies Choosing the right sequence based on relaxation times.
- FID / Pulse-Acquire: Essential for fast-decaying X-nuclei (e.g., ²³Na, ³¹P) with ultra-short T2*. Typically non-localized or uses 3D CSI (Chemical Shift Imaging).
- PRESS / STEAM: Standard for ¹H, but echo times are often too long for X-nuclei without significant signal loss.
- ISIS (Image-Selected in vivo Spectroscopy): Useful for ³¹P to achieve localized volumes before the FID without the T2* penalty of echoes.
Sequence Optimization Parameters needing adjustment for non-proton scans.
- Flip Angle (Ernst Angle): Must be optimized based on the specific T1 of the target metabolite, which often differs significantly from water/fat.
- TR (Repetition Time): Often shortened to maintain SNR over time, balanced against T1 saturation.
- Receiver Bandwidth (Hz/Px or total Hz): Larger chemical shift ranges for X-nuclei require higher bandwidths to avoid truncation and chemical shift displacement artifacts.

Overview of MRS Data Analysis

Unlike MRI, which reconstructs spatial images using frequency and phase gradients, Magnetic Resonance Spectroscopy (MRS) analyzes the chemical composition of a specific tissue volume by preserving the intrinsic frequency information (chemical shift). The primary goal of data analysis is to extract the concentrations of different metabolites from this frequency spectrum or its time-domain equivalent (the FID).

🧪 Try the Interactive MRS Analysis Demonstration

1. Pre-Processing

Cleaning the raw signal before fitting:

Coil combination: Optimally combining signal from phased array elements.
Frequency/Phase alignment: Correcting for scanner B0 drift during long acquisitions.
Filtration: Removing dominant residual signals (e.g., water or fat suppression) using techniques like HLSVD.
Eddy current correction: Using an unsuppressed water reference scan to fix phase distortions.

2. Quantification (Fitting)

Extracting metabolite concentrations:

Modeling: Constructing a mathematical model of the measured signal using known theoretical profiles (basis sets) or purely numerical methods.
Integration: The area under each spectral peak is directly proportional to the number of spins (concentration) of that metabolite.
Referencing: Converting arbitrary MR signal units to absolute concentrations (e.g., mmol/kg) using an internal reference (like water or total creatine) or an external phantom.

Data Analysis & Fitting Software

Time Domain Algorithms Fitting the FID signal directly (often preferred for X-nuclei).
- jMRUI (AMARES): A standard widely used for ³¹P and ¹³C data analysis. Excellent for interactive time-domain fitting and incorporating prior knowledge.
- FIDs & Non-linear Least Squares: Fitting exponential decays rather than Lorentzian lineshapes in frequency space handles truncation artifacts better.
Frequency Domain / Basis Set Linear Combination Fitting the frequency spectrum.
- LCModel: The historical commercial standard. Primarily ¹H, but can be adapted for X-nuclei if custom basis sets are simulated.
- FSL-MRS: A modern, open-source python toolbox. Highly flexible for simulating basis sets and fitting standard or X-nuclei data (runs locally or in Jupyter).
- Osprey: Open-source toolbox geared toward ¹H and edited-MRS, with automated pipelines.
- FID-A: An open-source MATLAB toolkit for simulating, processing, and analyzing in vivo MRS data.

Quality Assurance & Safety

Phantoms & Calibration Validation of system performance.
- Reference concentration phantoms (e.g., Sodium acetate for ¹³C, Phosphoric acid reference spheres for ³¹P).
- B0 field map stability and B1+ mapping (crucial for accurate flip angles with multi-nuclear surface coils).
Specific Absorption Rate (SAR) Safety concerns unique to decoupled spectroscopy.
- Continuous Wave (CW) or WALTZ broadband decoupling (delivering RF power to ¹H while detecting ¹³C or ³¹P) can rapidly approach hardware or regulatory SAR boundaries.
- Duty cycle monitoring and localized SAR models are necessary for FDA/IRB compliance.