Real-Time Smart Textile-Based System to Monitor Pressure Offloading of Diabetic Foot Ulcers

Abstract

Background:

The lifetime risk of developing a diabetic foot ulcer (DFU) is at least 25%. A DFU carries a 50% risk for infection and at least 20% of those receive some form of amputation. The most significant parameter that prevents or delays ulcer healing is high plantar pressure. To improve the patient’s healing process, the DFU’s plantar pressure should remain cumulatively low. Therefore, a tool that continuously measures the DFU loading, and provides real-time feedback can improve the healing outcome.

Methods:

We report the development of a system capable of continuously measuring the pressure, which could have applications to monitor DFU. The system contains a textile pressure sensor attached to a stretchable band, hardware that collects data and transmits them via Bluetooth to a phone, an app that gathers the data and stores them in the cloud, and a web dashboard that displays the data to the clinician. The sensor was characterized in vitro using the system, and the web-dashboard was developed and tested on simulated patient data.

Results:

We demonstrate the feasibility of developing the system and characterize the pressure response of the device. As a result, we demonstrate a viable method for monitoring DFU off-loading in real time.

Conclusions:

The presented study demonstrates the feasibility to develop a simple, modular wearable system that opens up new possibilities for diabetic foot ulcer care by providing a way of monitoring the pressure under the ulcer in real time.

The lifetime risk of developing a diabetic foot ulcer (DFU) may be as high as 34%.¹ A DFU carries a 50% risk for infection and at least 20% of those receive some form of amputation.^2-4 Foot ulcers manifest themselves as a region of tissue loss typically below the malleoli. Treatment of a DFU is essential to reduce the risk of complications and amputations. A mainstay of all DFU therapy is mechanical off-loading to relieve pressure and stress on the affected foot. Experts highlight the importance of patient adherence to off-loading in the healing of diabetic foot ulcers.⁵ The use of mobile technologies to support health behavior interventions (mHealth) is expanding rapidly as a means to improve monitoring of chronic diseases and adherence to healthy behaviors. The use of mHealth to monitor DFU offloading and alerting, with bidirectional feedback between patients and providers has the potential to improve the outcome from DFU treatment.⁶ Although there are already several off-loading solutions, such as total contact casts and removable walkers, none of these has a system for monitoring pressure over time to evaluate the effectiveness of the healing intervention. A key objective is to not only record pressure data, but also to share these data with the treating clinician and patient. The goal of the study is to develop a system to continuously monitor the stress applied to the ulcerated area of patients with DFU. We detail the steps taken to develop the prototype and characterize the data obtained by the system. We also present a web dashboard interface of the working prototype.

The goal of this article was to present the results of a prototype from a new wearable system, which can be used to monitor DFU off-loading, provide real-time feedback to the patient to increase compliance, and deliver data to the medical provider to support patient counseling and management decisions. We expect this system will be used in the future in experimental trials to test the impact of the offloading support technology on DFU patient outcomes.

Methods

Design of the Wearable Device

A small textile pressure sensor was used to detect pressure measurements on a localized area of the foot. Specifically, we used the textile pressure sensor patented (Patent Number: US 8,925,392 B2) and manufactured by Sensoria Inc, which is well-suited for this application due to its small size (a few hundred µm thick and a surface area of 6cm²) and its textile-based material (stretchable, bendable, washable, and sewable). The sensor behaves as a variable resistor (piezo-resistive), where the resistance is inversely proportional to the pressure. The sensor is illustrated in Figure 1, and it has an “S” shape.

Figure 1.

Textile pressure sensor. The sensitive area is 6 cm².

The sensor was sewn to the central part of an elastic band, which adhered to the site of the ulcer. The leads are represented by two conductive wires sewn directly with a sewing machine onto the band. The leads possess a wavy shape to allow the ability to stretch without breaking. Two metallic male snaps were positioned at each end of the lead. The snaps were the connection between the band and the electronic device, known as the Sensoria Core (SC) that will be applied to it.

The SC is a small device that collects the data from the textile pressure sensor. It has an analog front end that amplifies and filters the raw data, as well as an analog-to-digital converter (ADC). The data are then transmitted wirelessly via Bluetooth® Smart to a connected device using the internal Bluetooth module. The SC will provide real-time feedback to the patient via vibrations, noise and/or notifications in the phone app. This feedback is key to maintaining adherence to offloading. Figure 2 presents a picture of the prototype, including the band with the textile pressure sensor, the SC hardware, as well as a model of a foot showing a wrap around a foot ulcer.

Figure 2.

Prototype system, which presents a band with a sensor, an SC, and a foot mold.

The sensor response curve to known amounts of pressure between 0 and 845.5kPa was determined when increasing and decreasing the load. Fifteen sensors were tested to assess repeatability.

To apply pressure on the sensor we used a configuration where metal weights (with known) values were positioned above the sensor using a jig that covered only the surface area of the sensor. In this manner we guarantee that the pressure of the weight is only applied on the sensor. Using a scale we validated the repeatability of this configuration and determined that it can reliably apply pressure within a 5% error.

Design of the Web Dashboard

A web dashboard has been designed as an interface that will allow medical providers to review patient off-loading data. There are three components to the web dashboard, as shown in Figure 3. The first one (Patient Setup) enables the doctor to add new patients, change the data of existing patients, and adapt the pressure thresholds for each individual during the monitoring period. The second one (Clinician Dashboard) shows a summary from all patients. It presents a “traffic light” system highlighting patients in green (patients with adequate offloading), yellow (patients with marginal offloading), and red (patients with inadequate offloading who require immediate follow-up). Patients highlighted in red rise to the top of the list to facilitate quick review by the clinician. Each entry can be expanded to provide the clinician with more information about the selected patient. The third section (Patient Data) represents the screen where the doctor can visualize key metrics specific to an individual patient. The graph on the bottom has 24 bars that represent the hourly averages over the course of the day. The bars are interactive, and clicking on one of them will show the corresponding details for that hour.

Figure 3.

Main screen of the web dashboard.

The data stored in the cloud database are HIPAA-compliant to safeguard patient privacy.

Simulated patient data were included in the system to test the end to end web dashboard.

Results

The pressure response of the system was tested by applying and removing known amounts of weight over the whole surface area of the sensor. The measured data are presented in Table 1, and shown in the graph in Figure 4. The output of the system is in ADC values, a value that is proportional to the amount of pressure applied.

Table 1.

Measure ADC Values as a Function of the Pressure Applied (kPa) on a 6cm² Sensor.

Applied pressure (kPa)	ADC uphill	ADC downhill	Delta ADC
0	390	403	13
29.7	404	445	41
41.1	409	446	37
62.4	417	449	32
66.7	419	454	35
85.3	424	455	31
103.8	428	456	28
215.1	448	487	39
289.2	465	489	24
400.5	478	525	47
585.9	504	540	35
771.3	529	547	18
845.5	548	548	0

Figure 4.

Graph of the measured pressure values. The increase of pressure is shown in blue, and the decrease of pressure is shown in orange.

In Figure 4 we observe hysteresis which is due to the sensor substrate (eg, textile). The textile has mechanical properties that are different when the fabric is compressed and decompressed, thereby affecting the reading of the sensor resistivity. After repeating the experiment 15 times, the sensor to sensor variability was 22.93%. Although the error is high, the data are sufficiently accurate for the purpose of this application, where we are using the sensor as a detector of the presence or absence of offloading.

The web dashboard has been tested by simulating the presence of patients, registered in the Patient Setup section (Figure 5). For each patient we generated a simulation of pressure data within a 24-hour window. The summary of these data is observed on the Clinician Dashboard section and highlighted according to the traffic light rule previously mentioned. Figures 5, 6, and 7 show the Patient Setup, Clinician Dashboard, and Patient Data, respectively, which were obtained during the simulation.

Figure 5.

Patient Setup section.

Figure 6.

Clinician Web Dashboard section.

Figure 7.

Patient Data section.

Discussion

We present a fully functional prototype system designed to potentially aid in the treatment of DFUs by monitoring off-loading and communicating the results seamlessly between patient and clinician. The patient will be monitored continuously during the period of care and the doctor will be able to track the pressure values of each patient with this device to identify and address ulcers that are unlikely to heal.

With regard to the pressure measurements carried out with the band, which hosts the sensor, it is important to observe that the sensor has a slight hysteresis, as shown in Figure 4. However, the pressure detection is done in a precise manner. The SC allows the system to obtain data in real time. Furthermore, the data could allow the clinician and patient to work together to maximize activity whilst minimizing peri-ulcer damage during standing and walking.⁷

We have developed a web-based dashboard system that displays data in an orderly and intuitive manner. The “traffic light list” can be seen clearly in the Clinician Dashboard section and graphs generated by the Patient Data section facilitate and simplify the reading of the pressure data of each patient. We believe data visualization methodologies such as this may prove important to allow rapid assessment and understanding of key components of gait (such as cumulative pressure or dangerous peak pressure) for both the clinician and patient.

A main limitation of this system is that it requires the calibration of the sensors. Also, the sensors can degrade due to wear and tear; however, we expect that these sensors will be disposable and replaced on a frequent basis, perhaps weekly, or as dictated by wound care dressing changes. The system will need to be tested in the manufacturing process to guarantee proper behavior and high quality.

In this article we provide the proof of concept of a system that is capable of tracking changes in pressure using a textile senor integrated to hardware device that can provide patient feedback in real time as well as display results to medical providers via a web dashboard. Future studies will use this system in vivo on DFU patients to properly monitor and support adherence to offloading with the ultimate aim of improving outcomes.

Conclusions

The system described in this article marks an important innovation in the field of mHealth for DFU care. The ability to remotely monitor pressure in the area of an ulcer presents the opportunity to verify and encourage optimal off-loading in the community setting. The doctor will have the convenience of being able to track several patients at once within a single web dashboard and be able to quickly pinpoint the higher risk patients with inadequate off-loading.