Evaluation of optimal echo time for 1H-spectroscopic imaging of brain tumors at 3 Tesla
Elke Hattingen1, Ulrich Pilatus, Kea Franz
1Institute of Neuroradiology, Johann Wolfgang Goethe University of Frankfurt/Main, Frankfurt, Germany. elke.hattingen@kgu.de
Purpose:
To compare the spectral quality of short echo time (TE) MR spectroscopic imaging (MRSI, TE = 30 msec) with long-TE MRSI (TE = 144 msec) at 3 Tesla in normal brain and tumor tissue.
Materials And Methods:
Spectroscopic imaging (chemical-shift imaging (CSI)) data of 32 patients with histopathological confirmed brain lesions were acquired at 3 Tesla (3T) using TEs of 30 msec and 144 msec. Tumor-relevant metabolites (trimethylamine (TMA), creatine compounds (tCr), and N-acetylated compounds (tNAA)) were analyzed with LCModel software, which applies prior knowledge by performing a frequency domain fit using a linear combination of model spectra.
Results:
Short-TE spectra provided up to twice the signal-to-noise ratio (SNR) compared to TE = 144 msec. The estimated fitting error was improved up to 30% for TMA and tCr, but was slightly reduced (10%) for tNAA. Quantification in terms of absolute concentrations was consistent at both TEs.
Conclusion:
Since other metabolites observable at TE < 30 msec may be of diagnostic relevance, short-TE MRSI should be the preferred method at 3T for the evaluation of focal lesions in brain tissue; however, TE = 144 msec can serve as an option for MRS in regions with potential baseline problems.
Insights
Short echo time (TE) MR spectroscopic imaging (MRSI) at 3 Tesla offers superior signal-to-noise ratio and improved fitting accuracy for brain tumor metabolites compared to long-TE MRSI.
Area of Science:
- Neuroimaging
- Magnetic Resonance Spectroscopy
- Biomedical Engineering
Background:
- Magnetic Resonance Spectroscopic Imaging (MRSI) is crucial for non-invasively assessing brain tissue composition.
- Optimizing MRSI parameters like echo time (TE) is essential for enhancing diagnostic accuracy.
- Previous studies have explored various TE values, but direct comparison at 3 Tesla for focal brain lesions is needed.
Purpose of the Study:
- To compare the spectral quality of short-TE (30 msec) and long-TE (144 msec) MRSI at 3 Tesla.
- To evaluate the performance of these TE values in normal brain and tumor tissues.
- To determine the optimal TE for analyzing key brain metabolites.
Main Methods:
- 32 patients with brain lesions underwent 3 Tesla MRSI using chemical-shift imaging (CSI) at TE = 30 msec and TE = 144 msec.
- Tumor-relevant metabolites including trimethylamine (TMA), creatine compounds (tCr), and N-acetylated compounds (tNAA) were quantified.
- LCModel software was used for spectral analysis, employing prior knowledge fitting.
Main Results:
- Short-TE MRSI demonstrated up to double the signal-to-noise ratio (SNR) compared to long-TE MRSI.
- Fitting error was reduced by up to 30% for TMA and tCr with short-TE.
- Absolute metabolite concentrations were consistent across both TE values, indicating reliable quantification.
Conclusions:
- Short-TE MRSI (30 msec) is recommended as the preferred method at 3 Tesla for evaluating focal brain lesions due to superior spectral quality.
- Long-TE MRSI (144 msec) may be a viable alternative in specific cases with potential baseline instability.
- Further investigation into metabolites detectable at very short TEs (< 30 msec) could yield additional diagnostic insights.
