A Visual Servo Controlled Robotic System for MRI-guided Breast Biopsy

Achieving high repetitive accuracy and operation uniformity of small lesion targeting are the main challenges in robot-assisted procedures of MRI-guided breast biopsy. In this paper, we present a visual servo controlled robotic system for MRI-guided breast intervention. This robot system is developed based on the manual control version of previous publication. Zhang et al. (2019). The robot system exploits a cable-driven series-parallel mechanism and pneumatic controlled acupuncture mechanism, robot motion is guided by a machine vision system. The kinematic model of the robot is established, a control method of combining forward kinematics and visual servo control of a series-parallel robot is proposed to fulfil accurate control of the needle. A robot motion compensation method based on vision is developed, and the accuracy of the lesion targeting of the robot system is improved by establishing a feedback control mechanism. Repetitive accuracy results shows that the average error of targeting different lesions is between 1.29 mm to 2.42 mm, compared to previous manual control method, the visual control can provide significant improved accuracy. Comprehensive image quality test including signal-to-noise ratio (SNR) test, distortion test and uniformity test are carried out. The SNR variation is less than 10%, distortion is less than 1%, and uniformity is larger than 96%. Comparative experiments are conducted to verify the system’s visual servo control accuracy, MR image quality is evaluated to verify MRI compatibility under working condition. The experimental results show the robot satisfies the operational requirements and can provide effective assistance in clinical breast biopsy.