Increasing the Resolution of Clinical PET Systems by the Principle of Structured Illumination Microscopy

Y. Lang,C.-M. Kao,X. Chen,Q. Xie
DOI: https://doi.org/10.1109/nssmicrtsd49126.2023.10338700
2023-01-01
Abstract:We investigate applying the theory of super-resolution (SR) structured illumination microscopy (SIM) for increasing the resolution of clinical positron emission tomography (PET) imaging. The theory of SIM is based on modulating the image field to encode its high-frequency content above the bandwidth of the imaging device, and hence unobservable, into lower-frequency, detectable signals. Subsequently, high-resolution images are produced from a number of resolution-limited images that however encode the high-frequency content of the image field under examination by computationally restoring signals to their correct frequencies. In this work, we propose using a partially attenuating rotating ring, referred to as the modulator, in front of a stationary PET detector ring to create intensity modulation of the coincidence photons before detection. Using computer simulation data, we demonstrate a proof-of-concept for this SIM-inspired SR PET: We show that images reconstructed from multiple noise-free projection data obtained with the modulator at various rotational positions and a PET system made of 576 4.2 mm-width crystals can resolve 0.9 mm diameter sources. In practice, the characteristics of PET imaging physics and data noise will limit resolution recovery, but we also obtain promising initial results when using the ordered-subset expectation-maximization algorithm that is known to be relatively susceptible to noise. Further systematic studies are needed to better understand the theoretical aspects of the proposed method but we believe it can be practical after optimization of the design of the modulator and reconstruction method.
What problem does this paper attempt to address?