Super-resolution positron emission tomography by intensity modulation: Proof of concept.

We investigate a new approach for increasing the resolution of clinical positron emission tomography (PET). It is inspired by the method of super-resolution (SR) structured illumination microscopy (SIM) for overcoming the intrinsic resolution limit in microscopy due to diffraction of light. For implementing the key idea underlying SIM, we propose using a rotating intensity modulator of the radiation beams in front of the stationary PET detector ring to masquerade above-the-bandwidth signals of the projection data into detectable, lower-frequency ones. Then, an SR image whose resolution is above the system's bandwidth due to instrumentation is computed from several such measurements obtained at several rotational positions of the modulator. We formulated an imaging model that relates the SR image to the measurements and implemented an ordered-subsets expectation-maximization algorithm for solving the model. Based on simulation data produced by using an analytic projector, we showed that 0.9~mm sources can be resolved in the SR image obtained from noise-free data when using 4.2~mm detectors. Noisy data were produced either by adding Poisson noise to the noise-free data and by Monte-Carlo simulation. With noisy data, as expected, the SR performance is diminished but the results remain promising. In particular, 1.2~mm sources were resolvable, and the visibility and quantification of small sources are improved despite considerable sensitivity loss incurred by the modulator. Further studies are needed to better understand the theoretical aspects of the proposed approach and for optimizing the design of the modulator and the reconstruction algorithm in the presence of noise.