A 3D Structured Light Robot-Assisted System for CT-Guided Percutaneous Lung Targeting

OBJECTIVE: CT-guided percutaneous lung interventions remain challenging because respiratory motion and cumulative geometric errors can compromise targeting accuracy, procedural efficiency, and radiation burden. This pilot study aimed to develop and validate a robot-assisted framework for accurate, phase-locked lung targeting without continuous intraoperative radiation. METHODS: We developed a robot-assisted system integrating a 3D structured light scanner in an eye-in-hand configuration with a hybrid respiratory motion monitoring framework. The method combines single-acquisition 3D surface registration, high-frame-rate monocular Perspective-n-Point (PnP) tracking, and Kalman filtering to enable phase-locked instrument insertion. The system was evaluated across phantom, animal, and clinical cohorts and was compared clinically with conventional freehand CT-guided procedures. RESULTS: Comprehensive validation demonstrated preliminary evidence of sub-2 mm median registration accuracy and millimeter-scale targeting precision. In the clinical comparison, the proposed system reduced procedure time by 59% and significantly decreased radiation exposure relative to the conventional freehand workflow. CONCLUSION: By unifying calibration, registration, respiratory tracking, and robotic execution into a streamlined workflow, the proposed system achieved accurate phase-locked targeting while improving procedural efficiency in CT-guided pulmonary interventions. SIGNIFICANCE: This work provides a clinically viable robot-assisted solution for pulmonary interventions that simplifies the conventional calibration chain, mitigates line-of-sight and electromagnetic-interference limitations, and supports more standardized, efficient, and safer image-guided targeting.