3D pulse EPR imaging from sparse-view projections via constrained, total variation minimization.

Area of Science:

• Medical Imaging
• Radiation Oncology
• Biophysics

Background:

• Low tumor oxygen levels (hypoxia) correlate with resistance to radiation therapy.
• Tailoring radiation dose based on spatial oxygen distribution can improve treatment efficacy.
• Accurate imaging of tumor oxygenation is essential for guiding radiation oncology.

Purpose of the Study:

• To develop and evaluate an optimization-based image reconstruction algorithm for electron paramagnetic resonance imaging (EPRI).
• To reduce EPRI imaging time by minimizing projection scans while maintaining image quality.
• To improve the utility of EPRI for guiding radiation therapy by addressing artifacts from sparse-view data.

Main Methods:

• Developed a constrained, total variation (TV) minimization algorithm for image reconstruction.
• Algorithm enforces data consistency and minimizes total variation.
• Evaluated the algorithm using simulated, physical phantom, and pre-clinical EPRI data.

Main Results:

• The proposed TV minimization algorithm significantly reduces image artifacts compared to conventional filtered back projection (FBP).
• The algorithm demonstrates effectiveness in reconstructing images from sparse-view EPRI data.
• Both subjective and objective metrics showed the superiority of the TV algorithm.

Conclusions:

• The developed optimization-based reconstruction algorithm enhances EPRI image quality from sparse data.
• This technique offers a promising approach to reduce EPRI scan times.
• Improved EPRI imaging can better guide radiation therapy by accounting for tumor hypoxia.