Alignment of high resolution magic angle spinning magnetic resonance spectra using warping methods

Area of Science:

• Analytical Chemistry
• Spectroscopy
• Biophysics

Background:

• Magnetic resonance (MR) spectral peaks shift due to physiological and experimental variations.
• Accurate peak alignment is crucial for reliable multivariate analysis of MR spectra.
• High-resolution magic angle spinning (HR-MAS) MR spectroscopy generates complex spectral data requiring robust processing.

Purpose of the Study:

• To compare the performance of five spectral warping algorithms (icoshift, COW, fastpa, VPdtw, PTW) for aligning HR-MAS MR spectra.
• To evaluate the impact of peak alignment on data quality and subsequent multivariate analysis.
• To explore the potential of extracting diagnostic information from spectral peak shifts.

Main Methods:

• Comparison of five peak alignment algorithms on three sets of HR-MAS MR spectra with varying misalignments.
• Evaluation using quality criteria and Partial Least Squares Discriminant Analysis (PLS-DA).
• Analysis of spectral shift information from simulated and real MR data.

Main Results:

• Peak alignment significantly enhanced the internal similarity of the spectral datasets.
• The icoshift and COW algorithms demonstrated particular suitability for HR-MAS MR spectra alignment, likely due to their segment-wise approach.
• Successful discrimination between cancer and normal tissue samples was achieved using only spectral shift information.

Conclusions:

• Peak alignment is essential for processing HR-MAS MR spectra, improving data quality and analytical outcomes.
• The icoshift and COW algorithms are recommended for aligning HR-MAS MR spectra.
• Spectral shift information holds significant potential for biomarker discovery and disease diagnostics.