Origami robotics31 October 2017
Printed circuit boards for medical devices are a new technology that is redefining the world of microelectronics. We speak to Joshua Gafford, a robotics specialist from Harvard University, about how circuit boards could change the market for manufacturers.
The world of microelectronics might deal in the small, but it covers a huge area of medical device manufacturing. Microelectronic science is crossing over into the realm of 3D printing with the use of circuit boards for devices that can be used in any and all types of procedures.
One such device that is rapidly developing thanks to this technology is endoscopes, a market that has enjoyed rapid technological improvements over the past few years as the importance of the instruments has grown. Joshua Gafford, a PhD student at Harvard, has contributed to several studies on the topic, particularly around how endoscopes could move closer to other medical devices with integrated mechanisms, sensors and actuators via printed-circuit MEMS.
Gafford’s research focuses on developing smarter end-effectors and robotic systems for flexible and endoscope-based surgical systems. More specifically, he researches ways to develop mesoscale sensors, actuators, electronics and integration into robotic modules that can be mounted on the tip of commercially available endoscopes to provide greater distal dexterity and feedback sensing. “In doing so,” he says, “we hope to facilitate the process of using endoscopes to remove early stage gastric cancer in a minimally invasive way, by lowering the learning curve associated with performing complex therapeutic interventions endoscopically.”
Medical Device Developments: Would you be able to explain what exactly printed circuit boards do and what their uses are?
Joshua Gafford: Printed circuit boards (PCBs) mechanically and electrically support electronic components. They contain conductive lines, traces or pads that connect different components together electrically, allowing signal and power routing between devices. The solder used to mount components onto the board is electrically conductive and also serves as a strong mechanical adhesive. PCBs typically consist of a sandwich of different materials with various mechanical and electrical properties. Basic boards usually have an insulating substrate material where the substrate dictates the mechanical properties of the PCB; rigid substrates can be made from woven fibreglass (FR4) or phenolic paper, flexible substrates can be made from polyimide foils. There is a conducting material such as copper, a soldermask material that insulates copper traces from accidental contact with other conductive materials, and a silkscreen – a cosmetic layer that adds letters, numbers, symbols and graphics to the board. They can be single-sided, with all components mounted on one side; or double-sided, where components are mounted on both sides with conductive holes drilled through the board and copper-plated to transfer electrical signals from each side.
What new developments have you seen in the use of PCBs in medical devices over the past few years?
Most innovation in PCBs is on the integrated circuit side, with new semiconductor manufacturing techniques making smaller, lighter multipurpose electronic devices widely available and affordable. These innovations have substantially increased the achievable complexity of PCBs in smaller form factors.
Typically, PCBs in medicine have been used to drive electrosurgical generators, power endoscope vision systems, control RF ablation tools, power biomedical instrumentation and do a host of other things that require electronic power and control. All of this can happen outside and far away from the intervention site, meaning the electronics can be placed in big enclosures, and stored on shelves and racks in the operating room. However, active implantable and ingestible medical devices with integrated electronics not only have much stricter formfactor requirements, but also requirements with regard to biocompatibility, hermetic sealing, on-board power management and, in some cases, wireless communication.
As an example, consider the PillCam capsular endoscopy system from Medtronic. The PCB inside combines an on-board CMOS camera, LED illumination, wireless transmitter for transmitting images, battery and antenna, all in a tiny capsule.
Recent academic work has even looked at outfitting these pills with motors, improved sensing for localisation and mapping of the bowel, or retractable injection needles to enable localised therapy. We’re talking far-off ambitions. All these innovations are enabled by intelligent PCB design and advancements in semiconductor technology.
You contributed to a paper that says “recent advances in medical robotics have initiated a transition from rigid serial manipulators to flexible or continuum robots capable of navigating to confined anatomy within the body”. How would these robots and end-effectors benefit from PCBs?
Endoscope-based systems and continuum robots are dexterous machines capable of navigating the body’s intraluminal spaces. They enable a class of medical procedures called natural-orifice transluminal endoscopic surgery, where the body’s orifices are used as access points into the body.
These techniques promise improved patient outcomes by reducing morbidity and the associated reduction in rehabilitation time, and elimination of cosmetic scarring. However, endoscopebased procedures have historically been primarily diagnostic. This is due to factors such as limited functionality at the tips of these devices or limited dexterity; the inability to transmit forces due to the flexible nature of the tools; and a lack of feedback sensing that can lead to intraoperative complications. Simply put, we have the tools to navigate to the site of intervention, but the technology simply isn’t there to do meaningful things, because it is really hard to design sensors, actuators and electronics that are small enough to fit inside the body while also being biocompatible and robust.
In the landscape of manufacturing technologies, there exists a mesoscale gap; that is, a gap between millimetre and centimetre scales where no fabrication approach is really optimised for mass manufacturing of complicated 3D devices with integrated kinematics and electronics at this scale. At the higher end – the millimetre scale and larger – we have conventional manufacturing and assembly that can create discrete, heterogeneous components with tight tolerances. However, these discrete components must then be assembled, which drives up production costs. In addition, directly integrating electronics is not straightforward. At the lower end, silicon-based microfabrication can be used to build extremely small microelectromechanical devices in a planar fashion; however, developing 3D systems is prohibitively difficult.
Through my research, we are trying to address these problems through the fabrication of smarter end effectors that combine sensing, actuation and kinematics at millimetre scales. These devices will enable more dexterous endoscopic procedures through the addition of new degrees of freedom at the tip. A large part of my research focuses on the manufacturability of these devices and developing alternative means of cheaply developing mesoscale electromechanical systems. Specifically, we are tweaking the well-established printed-circuit manufacturing process to build active components with integrated electronics at millimetre scales in a way that is costeffective and scalable.
What future potential or breakthroughs do you see for PCBs in the world of medical devices?
PCBs have been mechanically passive, for the most part. Their function has been to mechanically ground electronic components, and route power and signals in between them. While flexible circuitry has been developed for signal routing on complex or non-planar surfaces, their use is primarily for passive wiring and there haven’t really been any efforts at actually exploiting this flexibility to build mechanically active circuit boards capable of moving and transforming.\
In applications where space is limited, innovations could make PCBs active devices as well, through the integration of kinematics and articulating structures. Such innovations may draw from the worlds of material science to build flexible circuits capable of withstanding substantial strains over thousands of cycles; origami-inspired engineering to leverage fold patterns to assemble two-dimensional PCBs into threedimensional functioning forms; electrical engineering to devise power electronics and signal conditioning methodologies in millimetre-scale packages; topology optimisation for board design and component layout in confined or limited form factors; thermodynamics for optimum heat dissipation, and RF engineering for wireless communications. In addition, academic interest in soft and origamiinspired, or self-folding, robots is steadily increasing, and a more active approach to PCB design and manufacturing could have a profound impact on those areas as well.
I see additive manufacturing potentially revolutionising the way PCBs are designed and fabricated. A company called Voxel8 has introduced a method for 3D printing electrically conductive traces in complicated 3D shapes. This could have substantial implications on how PCBs are designed and optimised when form factor or volumetric constraints exist.
What industrial or manufacturing benefits do you see PCBs offering the medical industry?
I’m not an expert on mass manufacturability of PCBs, so take this answer with a grain of salt. Printed circuit boards for medical products are inherently high risk, as their malfunctions could result in lost lives – especially for active implants – and lawsuits. As such, quality assurance is of utmost importance to ensure that the PCBs are manufactured to spec. Given this, while individual PCB design is by no means standardised, as the design varies depending on application and manufacturer, the manufacturing process is. As such, there are structures in place at the manufacturer level to ensure that the boards comply with strict QC protocols for medical-grade devices.
In addition, the printed circuit manufacturing process is extremely scalable, parallelisable and high throughput. It is straightforward to manufacture boards en masse in a cost-effective way.
Your own work covers modular robotic systems for therapeutic endoscopy and the use of PCBs. What does this entail and where could this technology be used?
Traditional applications of endoscopy have been primarily diagnostic in nature due to reasons I’ve described previously, such as limited distal dexterity and feedback sensing. A common thread is the difficulty associated with developing millimetrescale devices with integrated electronics, sensors, actuators and kinematics at millimetre-scales in a cost-effective, labourfree manner.
We are developing a process called printed-circuit MEMS, where we leverage PCB-inspired manufacturing techniques, as well as origami-inspired folding, to break PCBs out of the 2D plane and turn them into functional, articulating 3D mechanisms. We do this by leveraging material flexibility and mechanical pre-programming to create fold patterns that enable PCBs to assemble into complex 3D electromechanical devices through folding or self-assembly.
The purely planar fabrication process enables the use of high-tolerance laser micromachining, allowing us to build devices with features ranging from micro-to-meso scales in parallel. Origamiinspired folding techniques facilitate or even automate assembly. The original technique was devised to build microrobotic flying and crawling insects that were simply too difficult or timeconsuming to assemble by hand. I saw an opportunity to expand this work to the world of medical devices and robotics due to the inherent ability to fabricate complex, three-dimensional electromechanical devices at the mesoscale.
Rather than putting a printed circuit board into the handle of a tool, and using wires to transmit signals to the tool’s end effector that is itself manufactured using conventional machining approaches, what if you could turn the PCB itself into the end-effector? In doing so, you collapse several manufacturing processes such as conventional fabrication and assembly for the end effector components, PCB fabrication for the electronics, and wiring into a single monolithic process.