Optical Camera-Assisted Quality Assurance for Catheter Digitization In Ultrasound-Guided HDR Prostate Brachytherapy
Abstract
Purpose
Accurate catheter reconstruction is critical for treatment quality and patient safety in HDR prostate brachytherapy. Ultrasound guidance provides high-resolution imaging and excellent soft tissue contrast, yet manual digitization measurement under sterile conditions remains time-consuming and error prone. This feasibility study evaluates an optical camera-assisted approach, registered with a 3D model, to verify catheter positions and free lengths, aiming to enhance accuracy and streamline clinical workflow.
Methods
We implemented an optical camera-assisted verification system during HDR prostate procedures. Catheter positions and free lengths were manually recorded and modeled in 3D using scripting on an open-source platform. Concurrently, optical images were sequentially processed using Gaussian blur to reduce noise and highlight needle structures, followed by Canny edge detection and Hough line transformation with segmentation, enabling overlay with the 3D model for measurement verification.
Results
Sixteen catheters in one HDR Prostate treatment were analyzed. 3D model was generated within 1 minute using scripting. Grid template positions from optical images agreed with manual measurements for all 16 needles. Free-length verification was implemented with a linear fitting between the measured and those segmented from Hough line transformation, which showed a standard deviation of ±2.1 mm primarily due to catheter curvature and imaging noise.
Conclusion
Optical imaging integration with 3D model demonstrates a practical and touchless QA solution for HDR prostate brachytherapy under sterilization, improving digitization accuracy and reducing manual effort. This feasibility study lays the groundwork for automated, real-time verification systems to enhance patient safety and clinical efficiency.