Real-Time Image Reconstruction for Pulse EPR Oxygen Imaging Using a GPU and Lookup Table Parameter Fitting.

Area of Science:

• Medical Imaging
• Biophysics
• Physiology

Background:

• Tissue oxygenation is critical for understanding physiological and pathological processes.
• Electron Paramagnetic Resonance Imaging (EPRI) offers non-invasive 3D imaging of partial pressure of oxygen (pO2) with high resolution.
• Current EPRI image reconstruction is slow, hindering real-time applications.

Purpose of the Study:

• To develop and present methods for significantly accelerating EPRI image reconstruction.
• To enable real-time visualization of pO2 in vivo using EPRI.
• To improve the efficiency of EPRI for biological and medical research.

Main Methods:

• Implementation of graphics processing unit (GPU)-based 3D filtered back-projection for EPRI reconstruction.
• Utilizing lookup table parameter fitting to expedite the image reconstruction process.
• Combining GPU acceleration and lookup table fitting for enhanced reconstruction speed.

Main Results:

• Achieved acceleration factors exceeding 650 times compared to existing EPRI reconstruction methods.
• Demonstrated the capability for real-time reconstruction of EPRI pO2 images.
• Successfully visualized in vivo pO2 in tissues and tumors with high spatial resolution.

Conclusions:

• The developed methods enable rapid, real-time EPRI pO2 imaging in vivo.
• Accelerated reconstruction significantly enhances the utility of EPRI for studying tissue oxygenation.
• This advancement facilitates dynamic monitoring of pO2 in biological systems.