Magnetic resonance spectroscopic imaging reconstruction with deformable shape-intensity models

Area of Science:

• Magnetic Resonance Imaging
• Medical Imaging Analysis
• Spectroscopy

Background:

• Multidimensional magnetic resonance spectroscopic imaging (MRSI) data often suffer from low signal-to-noise ratio (SNR).
• Conventional methods like digital filters with signal apodization can degrade spectral lineshapes and linewidths.
• Improving SNR in MRSI is crucial for accurate diagnosis and research, particularly in neurodegenerative diseases.

Purpose of the Study:

• To introduce and evaluate a new deformable shape-intensity model (DSM) for enhancing MRSI data.
• To assess the effectiveness of DSM in improving SNR without negatively impacting spectral resolution.
• To compare DSM's performance against conventional apodization filters.

Main Methods:

• Development of a deformable shape-intensity model (DSM).
• Application of DSM to simulated and experimental in vivo (1)H MRSI data.
• Comparison of noise suppression and spectral quality between DSM and conventional apodization filters.
• Inclusion of data from cognitively normal (CN) elderly subjects and Alzheimer's disease (AD) patients.

Main Results:

• DSM demonstrated superior noise suppression in simulated MRSI data compared to apodization filters.
• Experimental data showed a 2.1-fold increase in SNR using DSM.
• DSM preserved spectral lineshapes and linewidths, maintaining spectral resolution.
• No distortion of spectral features was observed in the raw, unfiltered data after DSM application.

Conclusions:

• The deformable shape-intensity model (DSM) is an effective method for significantly increasing SNR in MRSI data.
• DSM provides superior noise reduction compared to conventional apodization techniques.
• DSM is recommended for reconstructing high-SNR MRSI data without compromising spectral integrity.