By Sarah L. Hemler, Sofia Lydia Ntella, Kenny Jeanmonod, Christian Köchli, Bhawnath Tiwari, Yoan Civet, Yves Perriard, Zoltan Pataky
A need exists for user-friendly, offloading footwear that reduces high plantar pressure to prevent and treat foot ulcers.
Figure 1. Schematic of the outer and inner (removable insole system) parts of the shoe.
Introduction
The high prevalence of lower extremity ulceration and amputation in people with diabetes is strongly linked to difficulties in achieving and maintaining a reduction of high plantar pressures (PPs), which remains an important risk factor. The effectiveness of current offloading footwear is opposed in part by poor patient adherence to these interventions, which have an impact on everyday living activities of patients. Moreover, the offloading devices currently available utilize primarily passive techniques, whereas PP distribution is a dynamically changing process with frequent shifts of high PP areas under different areas of the foot. Thus, there is a need for pressure offloading footwear capable of regularly and autonomously adapting to PPs of people with diabetes. The aim of this article is to present a review of the advancements toward this goal made by a multi-centered team from the Geneva University Hospitals (HUG), University of Geneva (UNIGE), and École Polytechnique Fédérale de Lausanne (EPFL). The team is developing intelligent offloading footwear that is designed to use a pressure feedback loop to automatically sense and redistribute PPs to prevent and treat diabetic foot ulcers.
The study authors are creating this intelligent footwear with an auto-contouring insole, which will continuously read PPs and adapt its shape in the forefoot and heel regions to redistribute high PP areas. The PP-redistribution process is to be performed consistently while the footwear is being worn. To improve adherence, the footwear is designed to resemble a conventional shoe worn by patients in everyday life. Preliminary pressure offloading and user perceptions assessments in people without and with diabetes, respectively, exhibit encouraging results for the future directions of the footwear. Overall, this intelligent footwear is designed to prevent and treat DFUs while enhancing patient usability for the ultimate prevention of lower limb amputations.
Intelligent Insole System
The intelligent footwear presented in this article consists of outer and inner (removable insole system) parts (Figure 1). Within the removable insole system, there is a pressure-sensing system coupled with miniaturized pressure-offloading modules. The system is designed to automatically detect the location of high PPs and correspondingly adjust the contour of the insole according to the user’s individual pressure needs (Figure 2).
Figure 2. (A) High pressure regions of interest and example high pressures on the plantar surface of the foot. (B) 2D and (C) 3D schematics of high plantar pressures (downward red arrows), which guide the automation and deformation of specific modules (in red).
Inside the footwear, there is a removable, intelligent insole system (Figure 1 – inner), which is made of several components working together to redistribute high PP. The system consists of a housing insole in which the pressure-offloading modules, batteries, and control electronics rest, and above which the comfort insole and pressure-sensing insole sit. The pressure-offloading modules operate independently and are connected to the control electronics via flex PCB through channels on the underside of the housing insole. Each pressure-offloading module consists of 3 primary parts: 1) Top – deformable bellow filled with magnetorheological (MR) fluid and top plug, 2) Middle – flow channels and valve, and 3) Bottom – deformable reflow membrane and auxiliary reservoir (Figure 3). The pressure-sensing insole consists of piezoresistive sensors (dynamic range: 0-800kPa; sampling frequency: 200 Hz) aligned directly over each corresponding pressure-offloading module. The thin-protective layer is the interface between the foot and the removable insole system with a goal of providing a moisture-absorbing and comfortable barrier between the mechanics and the foot.
The design of the removable insole system is such that when pressure is applied by the foot to the area above a module (eg, from standing or walking), the module will be triggered to operate in 1 of 2 states: 1) valve off, or 2) valve on. When the valve is off, the MR material remains in its fluid state and can move through the flow channels and the annular gap in the valve to be dispensed into the auxiliary reservoir (Figure 4A). The resultant movement of the MR material results in a maximum module compression of 2.5mm. When the force is removed, the reflow membrane forces the fluid to return into the deformable bellow above. However, when the valve is on, the exciting magnetic field (magnetic flux) causes the MR material to solidify in the valve channels and prevents the fluid from traveling to the auxiliary reservoir (Figure 4B). In the on state, there is a maximum module compression of 0.5 mm due to mechanical stabilization.
A baseline decision algorithm (based on the maximum peak pressure sensed above a module) will be employed to determine which modules will be turned off according to the user’s pressure needs. There will be a set number of modules that may be turned off at 1 time (Xred) to redistribute pressures. In the future, a trained and validated machine learning algorithm will be used to intelligently control which modules are off or on.
Figure 3. The pressure-offloading module frame and dimensions.
The design of the inner components is based on previous research and tailored to be suitable for footwear conditions. The size of the region above each module that can be deformed to achieve the intended pressure redistribution is a compromise between the complexity of the control system (in terms of both the hardware and the algorithm) and the accuracy in the determination of the magnitude and location of the peak PPs. In this respect, previous research underlined that in most cases, a sensor having a surface area of 1cm2 is sufficient (accuracy of ≈ 90%) to define the position and the proportions of the peaks of PPs. Thus, the reference value for the surface area exposed to loading (above each pressure-offloading module) is fixed to 1.6cm2, being a compromise between sufficient sensing/actuation resolution and system complexity. Each of the modules is waterproof and future designs will incorporate this same quality for all other electrical components. Further tests will also address safety features such as componentry heating and falls risk due to the elevated height of the footwear (measurement of ground to foot plantar height = 6.3cm).
The module’s performance and the system’s ability to reduce PP across the insole have been tested. In the module’s on state, it could sustain a load of 55N, which corresponds to 357kPa with a residual deformation of only 0.5mm. Thus, the performance of the module while on meets performance standards; with a PP of this magnitude, the module would likely be turned off to offload that region to prevent a foot ulcer. When the module was turned off during the tests, the module instantaneously deformed to 1.5mm which was the maximum deformation allowed for this test, and the force was reduced to 30N (corresponding to 214kPa). This final force was linked to the features of the deformable bellow and the hydraulic resistance. Furthermore, a preliminary walking study with 4 healthy, male adults was conducted to assess the PP reduction. Participants wore the prototype of the footwear with surrogate modules as they walked 10m at a comfortable walking speed. The deformation of the module between the on and off states allowed for a maximum reduction of PP of 18–24% directly over the module and 6–10% reduction in the area around the module when the peak starting PP ranged from 273–607kPa. Furthermore, for cases with an initial peak PP above 400kPa, there was a 20–32% reduction in peak PP.
Design for Patient Use and Adherence
Adherence is an essential aspect of medical device use. To understand the user perceptions and potential adherence barriers to this footwear, a pilot, in-person questionnaire and a larger, online questionnaire were conducted concerning the intelligent footwear presented in this article. Across the 2 questionnaires, people with diabetes (n=48), caregivers of people with diabetes (n=10), and healthcare professionals working with people with diabetes (n=65) from 30 countries on 6 continents gave important insight regarding the functionality, potential adherence, self-image, and aesthetics of the footwear. The questionnaires addressed the potential use and barriers to using this intelligent footwear based on previous work; questionnaires were administered and processed by the researchers with aid from clinicians. Generally, 95% of respondents thought that it would be beneficial to use the footwear and over 70% in each role stated that they would use the footwear or recommend it to their patients when available. Several parts of the questionnaire addressed self-image while wearing the footwear and perceived efficacy of using the footwear. The results informed aspects of this intelligent footwear design such as implementation of a sports shoe design which was the most preferred style among others. Future designs of this footwear will include styles of shoes for all occasions for men and women.
Figure 4. The valve off/valve on module design.
One of the limitations of other “smart” offloading footwear is the need for the patient to interact with a device and alter gait to relieve areas of high PPs. To lessen the required user-involvement and thus increase the likelihood of adherence, this intelligent footwear will have an autonomous, pressure-redistribution algorithm. As the individual wears the shoes, the insole will regularly read the PPs and automatically change the contour of the insole eliminating the need for the patient to interact directly with the device during the day. The only required interaction is the need to charge the shoes each day after wear. With a current of 0.7A, activating the on state of a module (200 ms) for 5,000 steps (recommended daily step count per foot for people with diabetes) would require 195mAh per module. One shoe of size EU 43 has 31 modules, which would require a total of ~6,000mAh. Therefore, a battery with 9,000mAh of energy (footwear has the capacity to house two batteries) is sufficient to provide a day’s worth of charge for each shoe. To apprehend complications with charging that could possibly reduce adherence, the footwear is designed to have a charging mechanism similar to technology that users may already operate (eg, cellphones, tablets). Furthermore, assessments of other adherence parameters have been performed and are ongoing in order to increase footwear adherence.
Conclusion
The presented intelligent footwear is designed to automatically and autonomously redistribute high PP under the feet of people with diabetes and specifically those with neuropathy. The mechanisms to offload the pressures use an intelligent, removable insole system which will actively adapt to the person’s foot while they are wearing the intelligent footwear. The footwear is designed to improve adherence through simplicity of user involvement and aesthetics resembling footwear not associated with a medical condition. Future versions will improve upon the technical and human factors aspects of the footwear to enhance flexibility, durability, battery life, usability, and aesthetics. The technological and adherence aspects of the footwear will continue to be tested and improved through clinical trials.
This article has been excerpted from “Intelligent plantar pressure offloading for the prevention of diabetic foot ulcers and amputations,” Front. Endocrinol. 14:1166513. doi: 10.3389/fendo.2023.1166513. Editing has occurred, including the renumbering or removal of tables and figures, and references have been removed for brevity. Use is per CC BY.
