A novel method using textural analysis of thermal images to predict the healing status of diabetic-related foot ulcers

This research plans to codesign and validate a fast, computer-aided, non-invasive test using thermal imaging that can accurately predict healing of diabetes related neuropathic ulcers at week 12 from an image taken at week 0 (baseline). Phase 1. Co-design process; there are 2 groups of participants: a) Individuals living with diabetes related foot ulcers b) Clinicians managing people living with diabetes who have a foot ulcer The 2 groups of participants meet on separate occasions. The co-design process involves small focus group discussions guided by an experienced facilitator with a pre-determined schedule designed to explore experiences and opinions on the topic and make use of interactional group exchange to gather data less accessible through individual interviews. We aim to use this approach to explore participants thoughts and address their needs regarding using technology to collect data to enable the prediction of DFU healing. We anticipate that each focus group meeting will be about 1.5 hours. The approximate timeline between each session will be between 4 to 6 weeks. The group discussions will run as follows: Session 1: Will include clinician participants and start with a discussion of the ‘why’ this device is being developed by the research team. The following will be then be raised for discussion by the participants – although the content will be shaped by the input provided by the participants, as they may raise aspects not considered by the research team: 1. What would the key drivers be for clinicians to want to use such a device?; 2. Specifics around the device itself: a. Range of costs they’d be prepared to pay for the device; b. What weight of the device they would be willing to accept; 3. Specifics around using the device will be drawn from using a persona – a person with a wound who is coming to the clinic for their first appointment, and an assessment will need to be undertaken. Clinician participants will work through this first appointment to unpack the following: a. The software of the device: i. Does it need to be connected to their Smartphone or organisation's database?; b. Specifics around the input/presentation of the information: i. What is the easiest way to enter the meta-data; ii. How should the information be presented ie. Binary – will heal/won’t heal? c. What are the critical aspects of the outcomes that will dictate your usage of the device? i. What would need to be included to guide the patient care plan? This information will be then taken by the researchers and the device will be shaped by this input. Session 2: Will involve clinician participants trialing the hardware to test it. Again, using a persona, clinician participants will be asked to consider: Weight, display, taking an image, entering meta-data, having the results displayed, and what staff need to do with the results so that it can be used in clinical care plans. If significant issues are identified, this may need to be repeated after adjustments have been made by the engineers in the research team. Session 3: Will involve people living with diabetes related foot ulcer participants, to provide their insights on what would be needed to make the device appropriate from their perspectives. Data obtained in these sessions will be used to adjust the prototype. This session will start with a discussion of the ‘why’ this device is being developed by the research team. The following will be then be raised for discussion by the participants, asking them to consider as a person who is undergoing an assessment by the clinicians at the clinic – although the content will be shaped by the content provided by the participants: 1. What are your thoughts about the device and what it is trying to achieve? 2. To test the usage of the device with participants – to invite feedback on the process, from their perspective; and 3. How should the information identified by the device be presented to them? Session 4: Will include clinician participants and the final version of the device will be generated including what people living with diabetes related foot ulcers have contributed. To review and finalise device prototype for the validation stage. We anticipate that the validation study will commence 4-6 weeks after the completion of the co-design (phase 1) study. Phase 2 - Validation study Consecutive individuals with diabetes, aged 18 and over, who attend the diabetic foot and podiatry inpatient and outpatient clinics at Austin Health, Melbourne, Australia will be invited to participate. Information will be collected from all participants at baseline, 1, 2 and 12 week follow up visits, and is expected to take between 10-15 minutes for each visit. Data collected during each visit include: * Thermal images of the foot ulcer; * Colour (RGB, red, green and blue) images of the ulcers; * Wound dimension (area) tracing and depth; and * Foot ulcer infection status (as per WIFI classification: 0 = no symptoms or signs of infection; 1: mild (less than or equal to 0.2cm cellulitis); 2: moderate (>2cm cellulitis/purulence); or severe (systemic response/sepsis) Thermal and RGB (Red Green Blue) colour images and a wound tracing will be taken of participant foot ulcers during routine dressing changes by one trained person to keep the data collection consistent. In addition, the following information will be collected: * Patient demographics: age, gender; * Past and current medical history, especially regarding their diagnosis and treatment of diabetes (eg type and duration of diabetes); * Comorbidities; * Their biochemical profile as clinically prescribed; * Current medication list; * The duration and description of the foot ulcer using SINBAD wound classification system characteristics (ie. Site, Ischaemia/Peripheral Arterial Disease (PAD), Neuropathy, Bacterial infection, Area, Depth); * Documented risk factors for delayed healing Thermal images (and therefore, the thermal imaging device) will only be used when taking thermal images at baseline, and then weeks 1, 2 and 12 (follow up visits).