Experimental evaluation of convolutional neural network-based inter-crystal scattering recovery for high-resolution PET detectors

Abstract

Objective: One major limiting factor for achieving high resolution of positron emission tomography (PET) is a Compton scattering of the photon within the crystal, also known as inter-crystal scattering (ICS). We proposed and evaluated a convolutional neural network (CNN) named ICS-Net to recover ICS in light-sharing detectors for real implementations preceded by simulations. ICS-Net was designed to estimate the first-interacted row or column individually from the 8×8 photosensor amplitudes.
Approach: We tested 8×8, 12×12, and 21×21 Lu2SiO5 arrays with pitches of 3.2, 2.1, and 1.2 mm, respectively. We first performed simulations to measure the accuracies and error distances, comparing the results to previously studied pencil-beam-based CNN to investigate the rationality of implementing fan-beam-based ICS-Net. For experimental implementation, the training dataset was prepared by obtaining coincidences between the targeted row or column of the detector and a slab crystal on a reference detector. ICS-Net was applied to the detector pair measurements with moving a point source from the edge to center using automated stage to evaluate their intrinsic resolutions. We finally assessed the spatial resolution of the PET ring.
Main results: The simulation results showed that ICS-Net improved the accuracy compared with the case without recovery, reducing the error distance. ICS-Net outperformed a pencil-beam CNN, which provided a rationale to implement a simplified fan-beam irradiation. With the experimentally trained ICS-Net, the degree of improvements in intrinsic resolutions were 20%, 31%, and 62% for the 8×8, 12×12, and 21×21 arrays, respectively. The impact was also shown in the ring acquisitions, achieving improvements of 11–46%, 33–50%, and 47–64% (values differed from the radial offset) in volume resolutions of 8×8, 12×12, and 21×21 arrays, respectively.
Significance: The experimental results demonstrate that ICS-Net can effectively improve the image quality of high-resolution PET using a small crystal pitch, requiring a simplified setup for training dataset acquisition.

Deep learning-based PET image denoising and reconstruction: a review

“…This synergy between state-of-the-art TOF and deep learning technologies has pushed the limits of TOF performance [ 244 – 246 ]. Undoubtedly, the integration of deep learning will play a pivotal role in enhancing the performance of not only PET imaging but also signal processing [ 247 – 250 ].…”

Resolving inter-crystal scatter in a light-sharing depth-encoding PET detector

“…We then quantified the FWHM of each count profile to estimate the contribution of the coincidence response function to the intrinsic spatial resolution of the detector under each positioning scheme (Moses 2011). The resultant improvement in intrinsic spatial resolution, compared to baseline centroid-based positioning, was quantified using the approach described by Lee and Lee (2023). Lastly, the fraction of coincidence events assigned to an incorrect LoR, termed 'Spillover', is also quantified for each positioning method.…”

A deep neural network for positioning and inter-crystal scatter identification in multiplexed PET detectors: a simulation study

“…Future work will extend this proof-of-concept algorithm to encompass DOI considerations, with implementation on a physical PET detector. True basis functions will be measured using coincidence acquisitions between the detector and a reference detector consisting of a single crystal needle, as in (Lee and Lee 2023). DNN training can be performed using a combination of simulated and real data.…”

“…In the case of light-sharing pixelated detectors, CNNs can improve event positioning through edge crystal identification (Labella et al 2020), DOI estimation (Petersen et al 2024) and ICS correction (Lee and Lee 2021, Petersen et al 2024. For example, Lee and Lee (2023) reported ∼40% accuracy in selecting the first crystal of interaction in ICS events and a resulting volumetric resolution improvement of 47%-64% in a detector with high light sharing ratio. However, all of the prior work assumes that each individual photodetector signal is accessible for analysis, which is not always the case.…”