THE PROCOMFORT KIT
Summary: For my senior undergraduate thesis at Harvard, I made 3D printed sensors that measure the stresses that form between a prosthesis-wearer's residual limb and her device; because the capacitive sensors are made of soft conductive polymer inks, they are durable, cheap, and comfortable to wear.
BACKGROUND
In the summer of 2017, I volunteered in prosthetic and orthotic clinic Prótesis Imbabura in Ibarra, Ecuador. When I wasn't working in the prosthetic fabrication lab -- reshaping plaster molds or thermoforming prosthetic sockets -- I sat in the waiting room talking with patients. Through these conversations, I learned that prosthetic legs are often extremely uncomfortable and painful. Many people showed me the bruises and pressure ulcers that had developed on their residual limb. Others said nothing but limped as they walked between the parallel bars with their newly-made prosthetic device. I later learned that this is a near-universal problem: around the world, approximately 25 percent of amputees experience such severe pain and discomfort that they stop wearing their prostheses.
THE PROBLEM
After diving into prior research on this area, I discovered this pain occurs when a prosthesis fails to protect sensitive regions of the residual limb -- such as bony prominences -- from rubbing and high pressures. More specifically, the pain experienced by prosthesis-wearers derive from shear and normal stress that are transferred from the prosthesis to the residual limb during walking, sitting or standing.
Digging into the source of this problem, I learned so many devices are painful because prosthetists simply don't have any tools to measure in-socket stresses and objectively assess prosthetic socket it. Instead, they rely solely on patient feedback—like where they feel pain and how that pain changes when they walk—to guess where a prosthesis is too tight or too loose.
After participating in a dozen prosthetic fitting sessions, I learned some of the reasons why this approach, which assumes a patient's reported pain is a proxy for prosthetic fit, often leads to ill-fitting prostheses.
1) Patients may underreport pain either due to diabetes-linked neuropathy or due to a fear that they’ll lose access to services if they complain.
2) Patients may not know what a prostheses should feel like, or, what is an unavoidable amount of pressure and what will eventually cause pressure ulcers. Imagine that you’re wearing ski boots for the first time – even when they fit well, they aren’t really comfortable. But if you're wearing ski boots for the first time, you may feel at a loss if the ski-boot-fitter asks "So do they fit?"
3) Many patients won't feel pain until after the prosthetic fitting session, given that pain and pressure ulcers often develop over time from repeated, constant prosthesis use. But by the time a patient feels this pain, they're often already back home it's too late to make changes to the prosthesis (for many people I spoke with, they had traveled 5+ hours by bus to get to the clinic).
Though patient feedback is and should remain a crucial part of the prosthetic fitting process, it is insufficient alone to guarantee that a prosthesis will provide long-term comfort and safety.
DESIGN AIM
Although in-socket stress is closely linked to prosthetic fit and comfort, there are currently no tools to help prosthetists quantify the stresses that occur at the residual limb-prosthesis interface. As a result, prosthetists must rely on inexact patient feedback alone to assess socket fit, causing many people with amputations receive ill-fitting and painful prostheses that they ultimately reject. After witnessing these shortcomings first-hand, I was inspired to create the ProComfort Kit: a tool to help prosthetists quantitatively assess prosthetic socket fit.
Given that patient feedback alone is insufficient to assess prosthesis fit, how could I empower prosthetists to create more comfortable prostheses for patients using stress sensors?
DESIGN APPROACH
During my time at Prótesis Imbabura, I talked with the prosthetists on staff and patients to understand what it would take to make a successful sensing system. I learned that the sensing system must do the following:
1) Measure physiologically relevant stresses.
According to prior research, the maximum stresses thought to occur between the residual limb and prosthetic socket are 80kPa shear and 350kPa normal stress. In order to make sure I capture the peak stresses on the residuum, I needed to make a sensor capable of measuring this full range.
2) Comfortable and unobtrusiveness when placed between the leg and socket.
Patients who I spoke with said that even a slight crease or bump in their liner or prosthetic sock can cause significant discomfort. Thus, the part of the sensing system that goes inside the liner should be no more than 1-2mm thick. And to avoid stress concentrations around the sensor when it's placed against the skin, the sensor's materials should be approximately as soft as the underlying leg muscles.
3) Easy to stick on the residual limb and stays in place during testing.
Because prosthetists only have a short time to test and correct a socket's fit, it was important the sensors could be placed on the leg rapidly, within 5-10 minutes. And to give accurate information about the location of high stresses, the sensors have to stay in place for 2-4 hours.
4) Be affordable for prosthetists and patients in the developing world.
After talking with Bob, the main prosthetist on staff at Prótesis Imbabura, he suggested he would only use such a system if it costs $500 or less -- about the price of one prosthetic device that he doesn't have to re-do if it's done better the first time.
Because capacitors generally exhibit higher measurement sensitivity and are less susceptible to environment-linked errors than FSRs or resistors, I decided to base my design off capacitive sensing. I developed a differential sensor design that would allow the sensor to measure both shear and normal stress simultaneously while maintaining a small overall size and profile. In terms of function, any time a normal and shear stresses impinges on a capacitive sensor, it deforms or compresses, generating a measurable change in capacitance between the common electrode and each of the 3 differential electrodes that can be parsed into x-y shear and normal stress.
DESIGN ITERATIONS
The first iteration of the Kit consisted of a soft differential capacitive sensor and data acquisition unit (DAU) capable of streaming real-time data to a nearby PC via Bluetooth. To ensure the sensor could be placed comfortably against the skin, I selected a soft, biocompatible thermoplastic urethane material (35A durometer) as the sensor's primary scaffold material. By 3D printing each layer of the capacitive sensor with fine tip nozzles, I ensured the sensors were thin and unobtrusive (~0.5mm), enabling it to be placed comfortably at the residual limb-prosthesis interface. Because these urethane materials are mechanically robust (3.5MPa tensile stress), the sensors are durable over repeat testing, never tearing or ripping. After characterizing the sensor, I verified that each sensor measures physiologically relevant stresses.
To understand how well the sensor worked in real-life applications, I traveled back to Prótesis Imbabura and conducted user testing. While there, I discovered several limitations of my design. First, in the first iteration of my design I used coaxial wires to shield sensors from noise and prevent signal saturation. However, during user testing these wires were bulky and limited the amputee’s range of motion. In addition, the rigid connectors interfacing the wires and sensors were fragile and caused discomfort on the residual limb. By connecting the sensor to the DAU through 3D printed traces inside the prosthesis, I ensured the final sensing system is fully soft inside the prosthesis and more comfortable.
In addition, when I tried sticking my sensors to the residual limb with KT tape and masking tape, I discovered it took a long time to place the sensors at the desired location and they didn't stay in place during testing. To make sure the sensor could be placed more easily on the limb, in the next iteration of the device I 3D printed the capacitive sensors directly on Tegaderm wound gauze, fusing the sensors to the Tegaderm's biocompatible adhesive backing. Thus, to place a sensor on the skin a prosthetist only has to peel off the wax-coated paper backing and seamlessly stick the sensor in place.
FINAL DESIGN
The final ProComfort Kit consists of soft 3D printed stress sensors that can be placed on a patient's residual limb before he dons his newly-made prosthesis. As the patient walks around with his new prosthesis, the sensors stream data real-time through the companion data acquisition unit to the prosthetist. This real-time data helps prosthetists identify regions of excessive stress on the residual limb and reshape prostheses to improve the overall stress distribution on the limb and fit. The result is safer, more comfortable prostheses.
