Optimization-based image reconstruction from sparsely sampled data in electron paramagnetic resonance imaging.

Area of Science:

• Medical Imaging
• Biophysics
• Computational Imaging

Background:

• Electron paramagnetic resonance imaging (EPRI) provides 3D spatial distribution of unpaired electron spin density, crucial for deriving tumor oxygen concentration images.
• Current EPRI reconstruction algorithms require extensive data, leading to long acquisition times and practical challenges, especially for in vivo animal studies.
• Acquiring data at sparsely sampled projection views can reduce scan time but often results in image artifacts with existing methods.

Purpose of the Study:

• To develop and investigate an optimization-based technique for reconstructing electron paramagnetic resonance (EPR) images from sparsely sampled projection views.
• To reduce data acquisition time in EPRI without compromising image accuracy.
• To address the limitations of existing algorithms in handling reduced data sampling.

Main Methods:

• Designed a convex optimization program to formulate the EPR image reconstruction as a solution.
• Tailored the Chambolle-Pock (CP) primal-dual algorithm to solve the convex optimization program.
• Validated the technique using computer-simulated EPRI data from numerical phantoms and real EPRI data from physical phantoms.

Main Results:

• The optimization-based technique successfully reconstructed accurate EPR images from sparsely sampled projection views.
• The developed method demonstrated the potential for significant reduction in EPRI data acquisition time.
• Computer simulations and physical phantom studies confirmed the efficacy and characterization of the technique.

Conclusions:

• The proposed optimization-based technique enables accurate EPR image reconstruction from limited, sparsely distributed projection views.
• This approach has the potential to significantly shorten EPRI scanning times, facilitating faster in vivo animal studies.
• The study highlights a promising method for overcoming practical challenges associated with prolonged data acquisition in EPRI.