A multiplexing method based on multidimensional readout method.

Abstract

Objective.To develop and validate a novel multidimensional readout method that significantly reduces the number of readout channels (NRC) in PET detectors while maintaining high spatial and energy performance.Approach.We arranged a3×3×4SiPM array in multiple dimensions and employed row/column/layer summation with a resistor-based splitting circuit. We then applied denoising methods to enhance the peak-to-valley ratio in the decoding map, ensuring accurate crystal-position determination. Additionally, we investigated the system's energy response at 511 keV and evaluated the suitability for both clinical and research PET systems.Main results.The proposed multidimensional readout method achieved a favorable multiplexing ratio, lowering the total NRCs without compromising energy resolution at 511 keV. Our tests demonstrated that a SiPM bias voltage of 31 V effectively balances gain and saturation effects, resulting in reliable energy measurements.Significance.By reducing system complexity, cost, and power consumption, the multidimensional readout method presents a practical alternative to conventional readout schemes for PET and other large-scale sensor arrays. Additionally, the approach can manage simultaneous multi-layer hits by arranging detector layers and, when needed, uses ICS detection to correct for scatter events. Its adaptable architecture allows scaling to higher dimensions for broader applications (e.g. SPECT, CT, LiDAR). These features make it a valuable contribution toward more efficient, high-performance imaging technologies in both clinical and industrial settings.