DIETI Contact person / Partner: Nicola Petra
Abstract
Computer assisted surgery has many advantages (minimized post-operative pain; sped up recovery; increased comfort for the surgeon; improved surgery's performance; reduced haemorrhaging; reduced risk of infections). There is however still a long way to go towards partial or full automation surgical interventions while the actual research activity is based on automating basic surgery.
The research activity will develop the hardware accelerators and the interfaces that will improve speed and accuracy of the atomic surgery tasks devised by other research groups that are part of the project. These tasks are based on the video elaboration of the Da Vinci stereo camera (the Da Vinci robot is available for the research group). Processing intensive tasks for video processing, feature extraction and tracking, will be leveraged through the use of high performance dedicated circuits.
Objectives
The research activity will assist in the automation of atomic surgical actions that joint together form a complete surgery activity. As an example, the actions of cutting, suturing, tissue repositioning, bone drilling, and tool exchange.
The project aims at developing:
- real-time visual sensing and elaboration of the scene;
- to track laparoscopic instruments;
- to automatically move the camera with no explicit human direction;
- to start performing simple autonomous tasks;
The considered test scenario will be the urological surgeries.
The research activity is expected to deploy hardware accelerators implemented on FPGA to assist the control and tracking algorithms. The initial steps will be the extraction of stereo data from the cameras (not an easy task due to the reduced volume of the laparoscopic surgical field) and the track of the tools.
Download
Available
Can be assigned to one doctoral student of cycle XXXVI