University of California, San Diego
Mechanical and Aerospace Engineering
MAE 156B: Senior Design Project (Winter 2024)
Background
Lung cancer is by far the largest cause of cancer deaths worldwide. The standard procedure for taking biopsies of potentially cancerous lung nodules involves putting a patient inside a CT scanner for imaging, applying stickers to target regions, then taking them out of the scanner to perform procedures. This can often lead to inaccuracies and cause the procedure to fail if anything shifts around. A robotic system would be able to operate inside the CT scanner, allowing real-time imaging and enabling doctors to have pinpoint accuracy. The Advanced Controls and Robotics (ARC) lab at UC San Diego created a robot called the CT-Compatible Robot and Needle Emplacer (CRANE) pictured below. The CRANE has 8 degrees of freedom and an extended arm with CT scanner compatible materials, allowing for precision control by a surgeon to perform biopsies while inside the bore of an active CT scanner.
CRANE v1 prototype performing needle emplacement on a surgical dummy within the bore of a CT scanner
Final Design
The goal of this project is to design and build a fully-functional CRANE v3 prototype, bringing the project closer to clinical translation and real-world use. Functional changes include creating a movement system for the CRANE to use within the CT scanner, perfectly counterweighting that system, and standardizing all the motors and controllers used on the robot with off-the-shelf components for future ease of use. Reach goals include creating an advanced control system for more precise dexterous manipulation of the CRANE with the new movement system, and force sensing at the needle tip to make biopsies more precise and safe.
After extensive theoretical comparisons of different design solutions, the final design was decided to be a 6-bar linkage on which the CRANE v3 would be mounted. The 6-bar would trace the circular bore of the CT scanner allowing for a precise arc to be traced by the robot in which it would not be able to hit either the patient or the CT scanner. The remote center of motion (RCM) of the CT scanner is precisely coordinated by a 6 bar linkage (seen below) meaning that the robot could decrease the radius of its arc to more closely adhere to the size of a given patient.
6 Bar Linkage tracing the Remote Center of Motion (RCM). The white arrow shows how the primary motion joint rotates.
6 Bar Linkage Free Body Diagram
The motor driving the 6 bar linkage is a custom Maxon Motor with a with an EPOS4 ____ controller. While the torque requirements of the robot means slightly different motors for each degree of freedom, every controller has been changed to an EPOS4____ which has excellent documentation and allows for closed loop control.
The weight of the robot and linkage mechanism was a problem, so a counterweighting system had to be put into place. A simple torsional spring would not suffice since the load required a massive unsafe spring size, and the equation of motion for the linkage scales sinusoidally unlike the linear torsion spring. The solution is a pulley and tension spring mechanism which counterweights the 6 bar near perfectly at every point, pictured below. The counterweighting is only 'near perfect' because in theory, the distance between the string and its mounting point on the linkage should be zero, but in practice, the thickness of the pulley and the design of the system play a role. The graph of the torque on each component in the system and the error is shown below.
Torque on motor throughout robot motion. The maximum error between counterweight and load (i.e. the maximum non-zero load) is 4.9 N*m
In accordance with NIH medical standards, the robotic system must be safe to operate on a human. This means multiple safety precautions in place. The robot will not move faster than 5cm/s, a standard medical operational speed. The gear ratio on the external gearbox (OR MAYBE INTERNAL) is a 1 to 100 reduction in rpm. This high of a gear ratio means that, even if the system loses power, the 6 bar linkage will not fall down on either side of the CT scanner.
Results / Performance
To reduce risk in the final design, a half scale prototype was built using plywood. This prototype served as a test for the linkage design, a prototype for the counterweight system, and a testbed for the manufacturing procedures. Regarding linkage design, the half scale followed the calculated path and compacted down to a small size. The counterweight system allowed the wooden linkage to be counterweighted perfectly at nearly every point; the linkage was not counterweighted well just below 90 degrees because of the spacing between the pulleys. From a manufacturing standpoint, the wooden links affirmed press fit sizing, link motion, and assumptions made about the ease of machine time.
CRANE has been significantly improved upon. Cracked acrylic parts were replaced with 3D printed spacers, and heavy metal plates were removed, bringing the weight of the system down by 5 pounds. Additionally, the pulley drive that translated the robot laterally across the CT scanner was removed.
The final metal linkage/CRANE system
NARRATED VIDEO