In neurosurgery, an error of a few hundred microns can result in irreversible brain damage. The risk has led to growing focus on minimally invasive surgery.
Surgeons perform procedures using increasingly sophisticated robots rather than scalpels and lasers. In the EU-funded EDEN2020 project, for example, researchers are developing a robotically steered catheter that can precisely deliver anticancer drugs directly to a brain tumour in situ. The system is only as good as its ability to position the needle as commanded, however. When it came time to specify the motion controller and drives for the four-axis robotic system that steered the catheter, researchers wanted compact, low-EMI, high-performance components.
A steerable needle
The EDEN2020 system is based on the Programmable Bevel-tip Needle (PBN), a flexible needle that is capable of advancing through the brain along a precisely-defined route that minimises tissue damage. The goal is for the robotic motion system to reference preoperative MRI scans and intraoperative ultrasound imagery during the procedure to generate path commands. When the PBN arrives at the tumour, it delivers a chemotherapy payload to the tissue.
The PBN consists of four interlocking longitudinal plastic segments nested together, each of which incorporates a drug delivery channel. The channel also contains a fiber-optic cable used for shape sensing. Each segment is driven at its distal end by an ironless motor. By pushing one segment or another forward so that it slides over the others, the system can cause the tip of the needle to curve by a specific amount. This process enables the needle to be gently navigated through the structures of the brain to reach even deeply embedded tumours.
It is an ambitious program with high performance demands. The task of controlling the robotic, steerable catheter through the brain is no simple matter. The four-axis system requires high synchronisation and high precision, as any errors could cause irreparable harm. At the low-level control range, the system needs to operate with an accuracy of 10 μm.
Performance is only one part of the challenge. Space is always at a premium in an operating theatre. The components used in the PBN need to be as small as possible. They also need to be quiet, both in terms of audible noise and EMI. Operating rooms are packed with instrumentation. The robot cannot interfere with imagery or the signal of a crucial instrument.
Drives and controllers
The PBN features four motors, each of which requires a drive. In addition, the overall system requires a high-performance motion controller to perform path planning based on closed-loop feedback and input from the MRI and ultrasound units.
For a drive, the Imperial College team selected the Elmo Gold-Twitter (G-TWI) servo drive. Just 35 mm x 30 mm, the compact Gold Twitter drive is essential for minimising the overall footprint of the portable surgery station. In addition, the servo drive’s extremely low EMI, resulting from a highly efficient pulse-width modulation (PWM) switching process, proved vital in this critical medical application. In an environment in which safety is a primary concern, the Gold Twitter, the smallest STO (SIL-3) certified drive available on the market, offers a huge advantage. Guiding the PBN requires the system to analyse the MRI and ultrasound data, then independently drive the four segments of the needle to direct the payload to the tumour. This needs to happen accurately, precisely, reliably, and at very high speeds. For a controller solution, the Imperial College team chose the Platinum Maestro (P-MAS) multi-axis motion controller.
The Platinum Maestro incorporates a multi-core processor and advanced multi-axis features, making it effective for highly synchronised systems. It includes a library of motion algorithms to simplify the implementation and control of machines and robots that need to be both fast and accurate. The Platinum Maestro features enhanced fieldbus support, including EtherCAT, cycling at a rate of 250 μs in the Imperial project.
The Platinum Maestro includes a number of features for ease of use. “In addition to the core benefits of the Elmo Motion Control technology, a key factor in our choice of control solution for the project was the reduced development time,” said Eloise Matheson, PhD candidate at the Mechatronics in Medicine Laboratory.
