Robot-assisted minimally invasive photoacoustic imaging for monitoring liver ablation using diffusing fiber illumination.

Abstract

Significance: Accurate intraoperative assessment of ablation completeness in liver radiofrequency ablation (RFA) remains a clinical challenge, as conventional imaging lacks real-time capability to delineate necrotic boundaries. Incomplete ablation increases recurrence risk, underscoring the need for real-time, high-resolution imaging with functional tissue differentiation.

Aim: We propose a robot-assisted photoacoustic (PA) imaging system employing a customized diffusing optical fiber to improve intraoperative monitoring of liver RFA.

Approach: The system integrates circumferential wide-field illumination for enhanced tissue coverage with robotic automated 3D scanning and co-registered ultrasound. Spectroscopic PA imaging differentiates necrotic from viable tissue based on optical absorption, whereas a standard Hough transform algorithm suppresses fiber-induced artifacts. Validation was performed using ex vivo and cadaveric swine liver studies.

Results: In cadaveric studies, 3D lesion mapping showed necrotic zones of $7.55\times 5.37\times 7.42\mathrm{mm}$ , closely matching gross pathology measurements (7.93 mm average diameter), confirming system accuracy.

Conclusions: The proposed system enables accurate, real-time visualization of ablation lesions in situ, offering a clinically viable approach to improve treatment precision and reduce recurrence in liver RFA procedures.