FeetSee Thermal Images Collection Protocol in Diabetes Patients

This is a prospective observational study with a single cohort of approximately 100 adults with type 1 or type 2 diabetes and a recently healed plantar DFU, conducted at hospital sites in the United Kingdom and at Vilnius University Hospital Santaros Klinikos in Lithuania. Eligibility requires peripheral neuropathy (defined by abnormal 10 g monofilament and/or vibration perception threshold testing) and at least one palpable foot pulse. Key exclusions include active DFU, active Charcot neuro-osteoarthropathy, severe peripheral arterial disease (no palpable foot pulses), major lower-limb amputation above the metatarsophalangeal joints, and medical or cognitive conditions that would prevent safe use of the device.

After informed consent and baseline assessments, participants are trained to use the FeetSee thermal camera, stand, and smartphone/tablet app at home. They are asked to remove footwear and socks and keep their feet non-weight-bearing for about 10 minutes before each measurement, then acquire at least one bilateral plantar image per day. Study staff provide intensive technical support during the first week and then schedule telephone check-ins approximately every three weeks, with additional troubleshooting as needed. Adherence is monitored via device-logged usage; failure to use the device for 21 consecutive days is considered loss to follow-up.

Thermal images are pseudo-anonymised and transferred securely to cloud infrastructure, where the FeetSee algorithm uses an active-shape model to detect foot contours and compares mirrored plantar regions for temperature differences above a predefined threshold (e.g.,>2.2 °C), indicative of local inflammation. The analyses performed by the device manufacturer (Diabetis JSC) will use the resulting temperature maps and clinical outcomes to estimate the sensitivity and specificity of the current algorithm for predicting incident DFUs and to develop refined algorithms. All other clinical and questionnaire data are analysed by the academic research team.

Feasibility outcomes include recruitment rate, retention, frequency of device use, and the number and nature of technical support contacts. Acceptance of the technology by both patients and healthcare professionals is assessed with questionnaires based on the Unified Theory of Acceptance and Use of Technology (UTAUT), covering performance expectancy, effort expectancy, social influence, facilitating conditions, self-efficacy, attitudes toward FeetSee, technology anxiety, and perceptions of personal DFU risk. Safety outcomes include the collection and classification of any adverse events and the identification of any device-related problems.

Inclusion criteria

• Males/females aged over 18 years.
• Type 1 or type 2 diabetes.
• History of plantar diabetic foot ulcer(s).
• Foot ulcer healed for at least 6 weeks before entry into study, as determined by a specialist healthcare provider.
• Presence of diabetic peripheral neuropathy defined by either of the following:

• Inability to detect the monofilament (10g) applied at two foot sites, including the apex of the Hallux (ADA guidelines 2008 Diabetes Care 31(1)).
• A Vibration Perception Threshold (VPT) >20V, and/or by inability to detect application of the monofilament at the apex of the Hallux.
• Presence of at least one palpable foot pulse.

Exclusion criteria

• Active foot ulceration.
• Active Charcot neuro-osteoarthropathy.
• Severe vascular disease (complete absence of any foot pulses).
• Lower limb/foot amputation greater than metatarsophalangeal disarticulation.
• Active malignancy, immunosuppressive disease.
• Physical or mental condition(s) that limit the ability to follow instructions for the study, based on clinical judgment by the physician, such as inability to use the device without assistance.
• Current participation in another clinical investigation of a medical device that is contraindicated for either study procedure/outcome.