Multimodal CT-ultrasound image-guided robotic system for automated abdominal puncture.

Computed tomography (CT) and ultrasound provide complementary strengths for image-guided interventions, yet single-modality guidance cannot deliver high spatial resolution, soft-tissue contrast, and real-time feedback simultaneously. This study introduces a multimodal image-guided robotic system designed for precision abdominal puncture based on CT-ultrasound fusion, which aligns preoperative three-dimensional CT images with intraoperative two-dimensional ultrasound images. Optical tracking is adopted to extract CT slices that spatially correspond to the live ultrasound plane, and the spatial alignment between them is achieved via a mutual information algorithm. A compact three-degree-of-freedom puncture actuator is designed, involving entry-point positioning, orientation adjustment, and needle advancement. The actuator rigidly integrates a probe clamping mechanism to preserve image-actuator co-registration. Phantom validation proved anatomically consistent fused displays and reliable guidance of the proposed system, and the puncture trials targeting renal calyces yielded a mean positional error of 1.17 mm and a mean angular error of 0.43°. The compact puncture actuator reduces footprint in the operating room and simplifies the calibration and clinical workflow. The proposed design provides a practical route toward safer and more reproducible minimally invasive abdominal interventions.