Estimation of pO₂ histogram from a composite EPR Spectrum of multiple random implants

Electron paramagnetic resonance (EPR) spectroscopy using oxygen-sensing implants can provide reliable and repeated measurements of the partial pressure of oxygen (pO₂) over a period of months or longer; however, it does not provide accurate information about the distribution of tissue oxygenation. While EPR imaging has the capability to provide spatially resolved oxygen data, it is time-consuming and not optimized for discrete number of implants. Previous reports suggest multi-site algorithms, which would require either the implants to be aligned in a certain way so as to deconvolve them using a linear magnetic field gradient or sparse imaging of the implants from a small number of 3D projections. In this paper, we present a simpler and much faster method to estimate the pO₂ histogram from a composite, single-scan EPR spectrum measured without applying field gradients to separate the EPR signals from multiple randomly placed oxygen-sensing implants. The method is optimized for a discrete number of implants, validated using simulations, experimental phantoms and in animal models. The results established the composite spectral fitting algorithm as a reliable and robust tool for multi-site oximetry.