Diabetes damages blood vessels and nerves, particularly in the feet, and can lead to severe infections that are difficult to treat. About 25% of people with diabetes will experience a diabetic foot ulcer (DFU) in their lifetime [2-4]. When circulation is so poor that a foot ulcer fails to heal or when treatment fails to stop the spread of an infection, sepsis can result. In such cases, amputation is often necessary. Diabetes is the leading cause of lower-limb amputations; DFUs precede approximately 84% of nontraumatic major amputations among people with diabetes [5-7]. The rates of recurrence of foot ulcers are very high, being greater than 40% after one year and 60% within three years [2,8].

Charcot foot, also called Charcot arthropathy, is one of the most debilitating outcomes of diabetes [9]. The condition causes increased blood flow to the foot and increased bone resorption. Immediately keeping all weight off, or off‐loading, of incipient Charcot foot appears to minimize fractures and incapacitating deformities. However, there is a potential for delayed diagnosis and therapeutic intervention as plain x‐rays may not show fractures at the early stages [10].

Diabetic foot ulcers result in considerable cost to the health care system when immediate ulcer episodes, social services, home care, and subsequent ulcer episodes are taken into consideration. Patients with a DFU were seen by their outpatient health care provider about 14 times per year and were hospitalized about 1.5 times per year. The cost of care for these patients was substantial, at about US $33,000 for total reimbursement of all Medicare services per year [11].

The total direct and indirect estimated cost of diagnosed diabetes in the United States in 2012 was US $245 billion. After adjusting for age group and sex, average medical expenditures among people with diagnosed diabetes were about 2.3 times higher than expenditures for people without diabetes [12] .

The national inpatient and emergency department bill summed to US $8.78 billion per year, averaging US $115,957 per case for major amputations [13].

Screening patients with diabetes to identify those at risk for foot ulceration has been shown to be beneficial [3]. DFU rates can be reduced with screening and appropriate intervention [4,14].

Self-care is a critical factor in detecting early signs of ulcers and injury. However, visual inspection has limitations (eg, patients with obesity or visual impairment cannot see their feet easily); hence, it is not very effective to identify the early signs. A recent study using a remote foot-temperature-monitoring system showed the ability to detect 97% of nontraumatic DFUs five weeks before they presented to the participant and/or clinician [15,16]. Additional options for detecting ulcers early in order to treat and heal ulcer wounds successfully may help prevent lower-limb amputations.

In diabetic foot complications such as foot ulcers and osteomyelitis, elevated temperatures in regions of the foot have been shown to be a precursor for ulceration [17]. In Charcot foot cases, increased plantar foot temperature is observed and is strongly correlated with the location of arthropathy. Temperatures decreased in a predictable manner as acute arthropathy resolved [18,19].

Thus, skin temperature assessment in persons with diabetes is a valuable tool for assessing inflammation in diabetic feet, as well as its resolution [20,21]. Home temperature monitoring has been shown to be an effective approach as an early warning system, to provide patients with objective feedback so they can modify their activity and protect their foot before ulcers develop; such monitoring is included in the International Working Group on Diabetic Foot clinical practice guidelines [16,22].

A handheld, infrared, dermal thermometer was designed to take temperatures on the bottom of both feet at six different spots each morning and compare these temperatures from spot to spot. Temperature differences of 4°F (2.22°C) or higher observed at comparable spots between the feet serve as an early sign of DFUs [16,23-25]. However, this tool has shown limited adoption. A reported shortcoming is that the manual temperature measurement on specific spots on the foot is subjective: asymmetric analysis tended to find false abnormal areas when the left and right feet had different sizes and shapes.

Digital health is a vast and burgeoning field and spans several aspects of health management. With the advent of the Internet of Things and the Internet of Medical Things coupled with smart devices, the potential for improved home care for medical applications is fast becoming a reality. Such devices can facilitate the management of chronic conditions at home, including the effective and timely management of DFUs. Diabetic foot scanners and voice-enabled scales are in development [26]. A “smart mat” allows daily measurement of plantar temperature, compares the temperature profile of the two feet, and aims to identify regions with increased temperature, detecting potential ulcer formation at an early stage [15,27].

Innovation in wearables has led to the development of “smart socks,” with embedded sensors for measuring temperature and reporting increases. A recent report describes socks made entirely of optical fiber [28]. While laboratory testing verified the accuracy of the sensors within, the fragility of the optical fiber resulted in limited usability. Another report describes socks that measure temperature at 10-minute intervals and contain wires [29].

All the tools described above are designed to measure temperatures once a day or at long intervals, are for clinic use only, or include wired data transmission. Once-a-day measurements present a risk of giving false positives. Continuous monitoring allows the assessment of temperature over longer periods, taking into consideration varying levels of activity over time, and thus has a greater potential to report consistent and clinically relevant temperature increases [30]. Continuous temperature monitoring can reduce false positives and has the potential to further improve home care and early detection.

Here, we introduce wireless sensor-embedded socks, made of neurofabric textile with microsensors embedded directly into the fabric—for continuous temperature monitoring of the feet of people with diabetes—and wireless reporting. They are designed to be easy to use and are washable as well as reusable.

A pilot study was conducted to assess (1) how comfortable sensor-embedded socks were for daily use and (2) whether observed temperatures correlated with clinical observations. Illustrative cases are presented.

The socks are made of “smart textile”: textile with microsensors woven directly into the fabric (Siren Diabetic Socks, Neurofabric, Siren Care Inc, San Francisco, CA). These virtually invisible sensors are seamlessly integrated into the socks to monitor temperature changes on the bottom of the feet. The sensor-embedded socks are designed to be reusable and are machine washable and dryable.

The sensors embedded in the socks are connected to a small tag on the sock, which encases a microcontroller unit, battery, and Bluetooth chip (see Figure 1A). The six sensors take temperature measurements at 10-second intervals to track temperature increases at the bottom of the user's feet, specifically at the hallux; metatarsal points (MTPs) 1, 3, and 5; midfoot; and heel (see Figure 1B). The data are stored in the tag and sent via Bluetooth to the phone paired with each pair of socks.

The mobile phone app can be programmed to generate alerts when the user’s feet show temperature increases that could be a warning sign of a potential ulcer (see Figure 2). In this study, the mobile phone app displayed temperature readings to the user, but alerts were not generated.

Sensors were tested prior to and after incorporation in the socks using a high-precision, thermostatic water bath (Zhejiang Jinbo Electronic Co, Ltd, China) and verified with a 0.01°C high-precision mercury thermometer. The stand-alone sensors were tested at four temperatures: 20°C, 25°C, 37°C, and 45°C.

The sensors woven into the socks were tested in the range 20°C-40°C. Three pairs of socks were tested by immersion in the thermostatic water bath for 10 seconds. The temperature recorded in the sock tag was compared with the reference standard.

A single-site study was conducted under an Institutional Review Board-approved protocol to evaluate the usability of the smart socks for patients with diabetic peripheral neuropathy (DPN). Informed consent was obtained from all patients.

A total of 35 patients, 18 years of age and older, from a private clinical practice were enrolled into three groups based on patient-reported medical history and/or medical documents. The groups were as follows: (1) Group 1 included subjects with DPN and no previous history of ulcers (n=11), (2) Group 2 included subjects with DPN and a previous history of ulcers (n=13), and (3) Group 3 included subjects with DPN and a current preulcer as determined by the investigator (n=11).

Subjects participated in two clinic visits. In the first visit, screening procedures were conducted, which included the following: a general physical exam performed by a board-certified podiatrist, visual foot inspection, digital photographs of both feet, and medical history intake. Subjects were provided with the socks and were given an Android mobile phone with the app needed for temperature monitoring. The socks were wirelessly connected with the mobile phone via Bluetooth. The patients were instructed to wear the socks continuously for 6 hours, after which the socks could be removed. The data were streamed via Bluetooth directly to the Android app installed on the phone provided to the patient during the screening and initiation visit. All data were stored in the sock tag and sent via Bluetooth to the phone paired to the socks that were assigned to the enrolled patient.

At the second or end-of-study visit to the clinic—7 days plus or minus 2 days from the screening and initiation visit—the socks were returned to the investigator and the patient was examined for potential adverse reactions. An exit questionnaire was completed to obtain usability information from the patient on the comfortableness of the socks, the ease of Android app use, and the practicality of integrating this specific system into the patient’s everyday life. Upon exit from the trial, data were exported from the Android phone to a secure laptop for analysis. All data collected were deidentified and only subject numbers were used for the duration of the trial.

A few illustrative cases are shown below, one from each of the three study groups.

To our knowledge, this is the first study to introduce wireless continuous temperature monitoring of feet for daily and home use in patients with diabetes. The sensor-embedded socks introduced here contain microsensors embedded directly and seamlessly into the fabric and are designed to look and feel like any other garment. Particularly important for daily and home use, the socks are wireless. Wireless transmission of data is achieved via Bluetooth technology. Through the mobile app, wearers can view the current temperature as measured at six points on the user’s foot. While the app was not set up to generate alerts in this study, users can receive a notification, alert, or text message when a temperature increase is detected between contralateral positions.

The aim of this study was to assess whether these sensor-embedded socks can measure temperature accurately on a continuous basis, whether the temperature findings are consistent with clinical observations, and to obtain feedback on patient experience in using the socks. The temperature studies presented here show that the sensors used in the socks are reliable and accurate at detecting temperature.

In this pilot study of 35 patients, participants found the socks to be no different from standard socks in terms of wearability and reported feelings of comfort and safety. Patients found the app interface to be useful. Furthermore, as shown in the illustrative cases, the temperature differences between the two feet as recorded by the sensor-embedded socks were consistent with the clinical status of the patient.

An earlier report described socks made of optical fiber designed for the clinic environment [28]. A more recent design of smart socks has woven-in sensors and measures temperature at 10-minute intervals, but contains wires [29]. The sensor-embedded socks described in this study are designed for daily wear, both inside and outside the home; are wireless; are machine washable; and do not need to be recharged. Unlike garments made of optical fiber, these socks are made of smart textile, which is designed to be made on standard industrial equipment and can be used anywhere without assistance.

For patients with diabetes and neuropathy, continuous temperature monitoring for the feet now offers information that was not previously available or possible, as was the case with the introduction of continuous glucose-monitoring technology for blood glucose levels.

Static or once-a-day measurements can present a risk of reporting false positives. With continuous monitoring, algorithms can be designed to identify and filter out outliers in measurements spanning several hours and, thus, can potentially reduce false positives by taking into consideration trends over time instead of a single static threshold. As shown in Case Study 2, temperature measurements of the entire foot may be particularly beneficial to patients with Charcot arthropathy. For patients undergoing treatment for an existing injury as in Case Study 3, continuous temperature monitoring provides an objective method to identify injuries.

Patterns of temperature can be obtained via continuous temperature monitoring that are specific to individuals and, in the future, variations from a person’s typical pattern may trigger alerts, rather than a single one-size-fits-all temperature threshold for all individuals. Monitoring patients over time may reveal temporal changes in individual temperature patterns. With further research, algorithms can be developed to detect temperature differences within one foot, without the need for the contralateral foot. Advanced statistical pattern recognition analysis could be used to determine patterns indicative of diabetes-related foot complications.

In future iterations, sensor-embedded socks can be coupled with built-in activity tracking to improve adherence and monitor patient compliance: data from the socks can be used to monitor patient activity and determine whether the patient is compliant with set activity guidelines.

This unique new data stream opens up questions regarding the manner in which the results are best reported, on the content and frequency of notifications, whether preulcerative lesions can be prevented from developing into ulcers, and whether amputations can be reduced. The pilot study reported here was not statistically powered to assess the performance characteristics of this novel device.

Further studies are planned to address such questions, with patient follow-up to obtain data on correlations of the temperature findings with patient outcomes.

The strengths of the study are as follows. The sensor-embedded socks were found to work reliably and consistently. The temperature differences reported matched clinical observations. Importantly, the study confirmed that patients can use the socks as a part of their daily lives, within or outside the home. Furthermore, the automatic collection and analysis of the data remove the element of subjectivity from the measurements as currently exists in visual inspection [30].

This study was not without limitations. As it was a single-day study, the findings could not be correlated with longer-term outcomes. More research is needed to further understand data points in continuous temperature monitoring, including as it relates to patient activity and timely intervention. Socks with built-in activity tracking and monitoring are planned to reliably and accurately measure activity concurrent with temperature measurement to further reduce subjective reporting. Future studies will be statistically powered to collect and analyze temperatures and correlate the findings to patient outcomes.

In this study, we explored the first use of wireless continuous temperature monitoring for daily and home use in patients with diabetes and neuropathy. This noninvasive device designed to behave as a normal sock is the first of its kind to combine wireless continuous temperature monitoring into a wearable device. The socks appear to the wearers to be no different than standard socks. When used with the mobile app, the wearer is kept informed about temperature increases in one foot relative to the other. The socks can reliably and consistently collect temperature data from the wearer’s feet, which are consistent with clinical observations. Continuous temperature monitoring has emerged as a promising tool which could serve as an early warning system for the management of foot ulcers, Charcot foot, and reulceration.