Artificial intelligence for automated detection of diabetic foot ulcers: A real-world proof-of-concept clinical evaluation.

OBJECTIVE: Conduct a multicenter proof-of-concept clinical evaluation to assess the accuracy of an artificial intelligence system on a smartphone for automated detection of diabetic foot ulcers. METHODS: The evaluation was undertaken with patients with diabetes (n = 81) from September 2020 to January 2021. A total of 203 foot photographs were collected using a smartphone, analysed using the artificial intelligence system, and compared against expert clinician judgement, with 162 images showing at least one ulcer, and 41 showing no ulcer. Sensitivity and specificity of the system against clinician decisions was determined and inter- and intra-rater reliability analysed. RESULTS: Predictions/decisions made by the system showed excellent sensitivity (0.9157) and high specificity (0.8857). Merging of intersecting predictions improved specificity to 0.9243. High levels of inter- and intra-rater reliability for clinician agreement on the ability of the artificial intelligence system to detect diabetic foot ulcers was also demonstrated (Kα > 0.8000 for all studies, between and within raters). CONCLUSIONS: We demonstrate highly accurate automated diabetic foot ulcer detection using an artificial intelligence system with a low-end smartphone. This is the first key stage in the creation of a fully automated diabetic foot ulcer detection and monitoring system, with these findings underpinning medical device development.

Saving the foot: Simple orthopaedic surgical intervention demonstrates improved outcomes and reduced costs.

BACKGROUND: Forefoot ulceration in diabetes requires significant resources, with high cost and low rates of success. The authors present the results of tendon procedures (percutaneous toe tenotomy and percutaneous tendo-achilles lengthening) under local anaesthetic to adjust mechanics in patients with diabetic neuropathic forefoot ulceration. METHODS: Retrospective review of electronic patient record of 19 patients (22 feet) undergoing local anaesthetic tendon procedures between April 2019 and April 2021 with a 12 month follow up period. Size of ulcer, rate of ulcer healing, complication rates and ulcer recurrence were recorded and compared to a population of conservatively-managed patients (14 patients, 15 feet) treated prior to the introduction of tendon procedures. All clinical information obtained from electronic patient records. RESULTS: All patients undergoing tendon procedures achieved complete ulcer healing at a mean time of 3.3 weeks for toe tip ulcers (after toe tenotomy) and 4.5 weeks for metatarsal head ulcers (after Achilles lengthening). There were no admissions for diabetic foot sepsis, reduced recurrence, reduced amputation rates and no mortality. Of the conservatively managed cohort, only 3 of the 15 achieved ulcer resolution without recurrence within the 12 month study period. The cohort managed conservatively had an average cost of £ 9902 per patient, per annum. The intervention cost was £ 1211 per patient, saving an average of £ 8691 per patient, per annum with ulcer resolution (88 % reduction in costs). CONCLUSION: Significant patient benefit, reduction in resource use and cost saving was seen with this simple intervention, which merits full evaluation in a clinical trial. LEVEL OF EVIDENCE: Level-IV.

Implementation of a novel mHealth application for the management of people with diabetes and recently healed foot ulceration: A feasibility study.

Background: Diabetic foot ulcers (DFUs) cause significant morbidity and mortality. Faster referral to specialist clinics is associated with a reduced risk of severe DFUs. The INTELLIN® diabetes management platform is a novel mHealth application for the management of recently healed DFUs and other complications, promoting engagement and expediting self-referral. Methods: To determine the acceptability, suitability, and usability of the INTELLIN® platform, time until reulceration, site, ischemia, neuropathy, bacterial infection, and depth (SINBAD) score and incidence of self-referral for recurrence were assessed in patients from the Salford Royal NHS Foundation Trust. Patients and clinic staff also assessed platform usability. A Markov cost-utility model was used for the health economics and outcomes research analysis. Results: 197 patients were assessed for eligibility and 15 entered the full analysis set (FAS). Through Week 52, 8/15 patients experienced recurrence, with a mean SINBAD score of 2.1 and mean duration of 2.6 days. Mean time to recurrence was 273.0 days (95% confidence interval 74.0, 484.0). No patients self-referred. Initial qualitative data showed high platform usability. The INTELLIN® platform only required a relative reduction in recurrence of 5% versus standard of care (SoC) for an incremental cost effectiveness ratio of £20,000 per quality-adjusted life-year, suggesting potential for significant cost savings upon wider adoption. The barriers to enrollment encountered demonstrate the impact of socioeconomics on mHealth. Conclusions: These results suggest that the INTELLIN® platform is required to provide only a small reduction in recurrence compared to SoC to be a cost-effective strategy for prevention of recurrent DFUs.

Mortality in 98 type 1 diabetes mellitus and type 2 diabetes mellitus: Foot ulcer location is an independent risk determinant.

INTRODUCTION: We previously demonstrated in both a longitudinal study and in meta-analysis (pooled relative-risk RR, 2.45) that all-cause mortality is significantly higher in people with diabetes foot ulceration (DFU) than with those without a foot ulcer. In this prospective study, we looked at the factors linked to mortality after presentation to podiatry with DFU. METHODS: Ninety-eight individuals recruited consecutively from the Salford Royal Hospital Multidisciplinary Foot Clinic in Spring 2016 were followed up for up to 48 months. Data concerning health outcomes were extracted from the electronic patient record (EPR). RESULTS: Seventeen people (17) had type 1 diabetes mellitus, and 81 had type 2 diabetes mellitus. Thirty-one were women. The mean age (range) was 63.6 (28-90) years with maximum diabetes duration 45 years. Mean HbA1c was 72 (95% CI: 67-77) mmol/mol; 97% had neuropathy (International Working Group on the Diabetic Foot (IWGDF) monofilament); 62% had vascular insufficiency (Doppler studies); 69% of ulcers were forefoot, and 23% of ulcers were hind foot in location. Forty of 98 (40%) patients died in follow-up with 27% of death certificates including sepsis (not foot-related) and 35% renal failure as cause of death. Multivariate regression analysis indicated a 6.3 (95% CI: 3.9-8.1) fold increased risk of death with hind foot ulcer, independent of age/BMI/gender/HbA1c/eGFR/total cholesterol level. CONCLUSION: This prospective study has indicated a very high long-term mortality rate in individuals with DFU, greater for those with a hind foot ulcer and shown a close relation between risk of sepsis/renal failure and DFU mortality, highlighting again the importance of addressing all risk factors as soon as people present with a foot ulcer.

Evaluating the effect of a haemoglobin spray on size reduction in chronic DFUs: clinical outcomes at 12 weeks.

A recent multi-centre observational evaluation investigated the effect of a topical haemoglobin spray (Granulox, Infirst), used as an adjunct to standard care, on wound size reduction in 17 patients (4 females/13 males) with 20 chronic diabetic foot ulcers (DFUs) over a 4-week period. In 14 of the 18 wounds that completed the evaluation (one patient dropped out due to an infection) there was a mean reduction of 53.8% (range: 11.9-100%). The product was acceptable to both patients and clinicians, who all found it easy to use. This article describes the outcomes for the remaining 13 patients (with 15 wounds) who continued using the spray after the 4-week evaluation ended. (Data are not available for two patients and the one patient who healed during the 4-week evaluation.) By 12 weeks, three wounds (20%) had healed, eight (53%) were progressing towards healing, three (20%) increased in size and one (7%) was slow healing.

Evaluating the effect of a haemoglobin spray on size reduction in chronic DFUs.

AIM: The aim of this multi-centre observational evaluation was to assess the percentage reduction in wound area of non-healing diabetic foot ulcers (DFUs), treated with Granulox haemoglobin spray over a 4-week period. Secondary outcome parameters--for example, adverse events, patient acceptability and ease of use--were also recorded. METHOD: After a run-in-period (2 weeks for existing patients and 4 weeks for new patients) to determine if the wounds were non-healing despite receiving local best practice, patients whose foot ulcers had decreased in size by < 20% were then entered into the evaluation. A sample of 17 patients (4 females and 13 males), comprising 4 with type 1 and 13 with type 2 diabetes, with a total of 20 DFUs, met the inclusion criteria. These data were collected from six sites across the UK. RESULTS: There was an overall positive reduction in size in 15 of the wounds, equating to a mean reduction of 53.8% (standard deviation (SD): 26.6; range: 11.9-100%). One participant, with two ulcers, had to be withdrawn due to infection. All clinicians and participants found the product easy to use. CONCLUSION: The addition of a topical oxygenation therapy in this cohort of non-healing DFUs showed reduction in wound surface area and progression to healing. The product was also found to be acceptable and very easy to use by both participants and clinicians.

Debridement of diabetic foot wounds.

The management of foot ulcers in patients with diabetes is a complex and increasingly common problem. This article outlines the core principles of best practice for these patients focusing, in particular, on debridement. The article defines debridement, looks in detail at the rationale and evidence for its use in wound bed preparation and examines the various debridement methods available.