# Super-Resolution Reconstruction and Detector Geometric Error Correction for Parallel-Beam Low-Resolution Multi-Detector SPECT: A Proof of Concept

## TL;DR

This paper introduces a new method to improve SPECT imaging resolution using super-resolution and neural networks to correct detector errors.

## Contribution

A novel super-resolution reconstruction method with geometric error correction for multi-detector SPECT systems is proposed.

## Key findings

- The proposed method achieves a two-fold resolution improvement in SPECT imaging.
- Neural networks effectively estimate and correct geometric errors among low-resolution detectors.

## Abstract

Single-Photon Emission Computed Tomography (SPECT) is characterized by inherently low spatial resolution, which limits its ability to detect small lesions. To address this, we propose a novel super-resolution (SR) reconstruction method for a multi-detector, low-resolution (LR) SPECT system. Our approach integrates data from offset detectors and employs a neural network to correct for inter-detector geometric errors. Numerical simulations verify that this method effectively doubles the image resolution, presenting a promising solution for advancing SPECT imaging.

Objectives: Due to collimator limitations, Single-Photon Emission Computed Tomography (SPECT) suffers from relatively low spatial resolution, which hampers the detection of small lesions. This study proposes a super-resolution (SR) reconstruction algorithm for a parallel-beam, low-resolution (LR) multi-detector SPECT system and employs a neural network to estimate and correct for geometric errors in the LR detectors. Methods: A parallel-beam LR multi-detector SPECT system is presented, in which the detectors perform relative sub-pixel shifts. At each sampling angle, an SR reconstruction algorithm synthesizes high-resolution (HR) SPECT images from LR projections acquired by four offset LR detectors. To correct for geometric errors among these detectors, a randomly distributed gamma point source was designed to generate training data. A neural network was then employed to estimate the geometric errors, thereby refining the SR reconstruction. Results: Numerical simulation demonstrated that the proposed neural network could accurately identify the displacement-based geometric errors of the LR detectors. Utilizing these estimated parameters to correct the SR reconstruction process yielded results comparable to those obtained from direct reconstruction of HR projections, achieving a two-fold resolution improvement. Conclusions: Preliminary proof-of-principle for SR reconstruction in a parallel-beam LR multi-detector SPECT system was established. Further validation of the hardware performance is warranted.