Physics Informed Neural Network for Fitting Ultra High Field Magnetic Resonance Spectrum

Key Points

• The aim is to improve the fitting of magnetic resonance spectra using a physics-informed neural network.
• Utilized a physics-informed neural network for data fitting.
• Applied the model to ultra high field magnetic resonance spectra.
• Conducted performance comparisons with traditional fitting methods.
• Achieved enhanced accuracy in spectrum fitting.
• Reduced computational time compared to conventional techniques.
• Demonstrated robustness across varied magnetic resonance data sets.