ROBOT

The first year of postdoctoral was dedicated to implementing a visual servoing scheme based on the 3D imaging (C-scan) obtained from the OCT for guiding a robot during the intra-operative phase in order to perform a repeatable optical biopsy.
In the literature, the medical imaging (such as ultrasound, MRI, confocal microscopy, etc.) is used to synthesize different control laws which allow surgical tasks to be performed (e.g., biopsy, ablation, etc.). However, the OCT is not a common sensor in the visual servo loop despite its very interesting characteristics (e.g., real-time acquisition, 3D images, high resolution, etc.). Indeed, there are some works based on the usage of the wavelet transformation of the 2D OCT images (B-scan) in the control loop, but these works are limited to controlling only 3-DoF (Degrees Of Freedom) of the robot.
This research work aims to develop control methods, going beyond the state of the art, by allowing the 6-DoF of the robot to be controlled using the volume image (C-scan) obtained by the OCT as the control law input signal [1]. In fact, the PCA (Principal Component Analysis) technique is applied for estimating the pose of the C-scan with respect to the OCT. This information is used in the control loop for reducing the pose error between the actual one and the desired/reference one.

The second year was devoted to realize a final prototype of the project ROBOT which demonstrates the advancement done by the different teams. Indeed, the final demonstrator integrates a novel miniaturized OCT probe with an 8 mm diameter robotic endoscope. The micro-robotic system places accurately the OCT scanner with respect to suspicious tissue, and holds it stable with respect to physiological motion.