Development of a Portable PET System W/Real-Time Us Tracking for Interactive Point-of-Care Imaging

Le résumé fourni par la source

This study presents an updated Point-of-Care (PoC) PET system integrated with ultrasound (US), designed to enhance diagnostic imaging and image-guided interventions in environments where conventional PET/CT systems are impractical, such as operation room or intensive care units. We constructed a prototype dual-modality PoC-PET/US system using a Kuka KR 10 R1100 sixx robotic arm, a rotation stage, eight PET detector modules, and an Atracsys stereo camera (FusionTrack250) all mounted on a trolley. It employs list-mode data acquisition and GPU-based MLEM reconstruction engine for rapid on-the-fly image processing. A Python-based control interface streamlines robotic positioning, data acquisition, event sorting, and image reconstruction, reducing overall scanning and reconstruction time by 20 %. The PET imaging capability of the system was evaluated using a fillable tumor phantom with lesion sizes from 3.3 mm to 11.4 mm and a tumor-to-background ratio of 20:1. The PET/US co-registration capability was evaluated using a silicone phantom containing64to compare with conventional PET/CT. Lastly, PET images were collected for FDG and64injected mice to exhibit tracer uptake. Initial results indicate that PoC-PET images closely approximate the resolution of clinical PET/CT systems. Stereo camera accuracy was validated by mounting a passive marker on the robotic arm, yielding sub-millimeter translational and sub-degree rotational error relative to ground truth. The integration of PoC-PET with free-hand US tracking marks a significant advancement in dual-modal imaging, offering flexible, high-performance bedside diagnostics that complement traditional setups; ultimately enabling real-time image guided interventions such as biopsy and ablation procedures.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.