Design and performance evaluation of a GAGG:Ce long-bar Compton camera with dual-ended DOI readout using Geant4...

Compton cameras enable medium- and high-energy gamma-ray imaging without mechanical collimation. Extending an individual detector element increases the active volume without adding readout channels, but also increases axial position uncertainty and can degrade angular resolution. This study evaluated whether a cerium-doped gadolinium aluminum gallium garnet (GAGG:Ce) long-bar Compton camera with dual-ended depth-of-interaction (DOI) readout could mitigate this trade-off. A Geant4 detector-response model reconstructed deposited energy and axial interaction position from the two end signals. Simulated reference-bar energy and DOI resolutions were comparable in magnitude to published measurements. The proposed design was compared with a conventional two-layer planar configuration at matched crystal cross section, transverse element count, and total readout-channel count. The effects of bar length and detector gap on valid two-crystal Compton-event efficiency and the angular resolution measure (ARM) were investigated. Increasing bar length and reducing detector gap improved efficiency, although the gain diminished above approximately 75 mm and a shorter geometric baseline broadened the ARM. For an on-axis 511-keV pencil beam and the adopted event-selection criteria, the 75-mm-bar, 10-mm-gap design achieved a valid-event efficiency of 0.836% and a full width at half maximum (FWHM) of 11.42° for the fitted central ARM peak. The conventional reference at a 10-mm interlayer gap yielded 0.172% and 12.49°, respectively. The DOI configuration had a non-Gaussian ARM response; the practical procedure correctly ordered 60.4% of accepted validation events with an identifiable true order. The approximately 4.9-fold efficiency difference reflects both increased active volume and DOI-enabled axial localization.