Diabetic foot complications are a major cause of morbidity, disability, and amputation worldwide, yet many complications are preventable through effective self-care. Limited evidence is available on foot self-care knowledge and behaviours among adults with diabetes in Ras Al Khaimah, United Arab Emirates.

Haemodialysis patients face a markedly elevated risk of foot lesions because of coexisting diabetes mellitus (DM), peripheral arterial disease (PAD) and uremia-related impairment of wound healing. Structured outpatient strategies and the optimal place of basic fibroblast growth factor (bFGF) within a stepwise algorithm remain poorly defined. To report complete healing and its durability in haemodialysis patients treated with bFGF for foot lesions, and to propose a stepwise limb-salvage algorithm with prespecified escalation criteria in which bFGF precedes dehydrated human amnion/chorion membrane (dHACM) or surgery. Retrospective observational case series of 20 haemodialysis patients receiving bFGF (trafermin; Fiblast Spray) for foot lesions across three outpatient wound-care programmes. Wound status was documented by area, depth, tissue exposure and WIfI classification; perfusion was assessed in every patient by toe-brachial and ankle-brachial indices, duplex ultrasonography and angiography, independently of any pre-existing recorded diagnosis. Complete healing was defined a priori as 100% re-epithelialization confirmed on two consecutive visits at least 2 weeks apart. Tier entry criteria and escalation triggers were prespecified, using a wound-area reduction below 50% at 4 weeks as the principal threshold. The broader haemodialysis population (n = 212) served as contextual comparator. Mean age was 75.0 ± 7.3 years (range, 59-92); there were six men and 14 women. DM was present in 15 patients (75.0%) and registry-recorded PAD in 8 (40.0%), against 56.6% and 17.9% in the broader population; objective testing identified hemodynamically significant PAD in 12 patients (60.0%), all of whom were revascularized. Mean baseline wound area was 5.5 ± 3.4 cm (2), and WIfI stages 1-4 comprised 5, 8, 6 and 1 patients. Twelve patients (60.0%) were managed with bFGF alone, five (25.0%) required dHACM and three (15.0%) required surgery. Complete healing was achieved in 18 of 20 patients (90.0%) at a mean of 8.9 ± 4.4 weeks, with no amputation of any level. Over a median follow-up of 13.0 months (range, 6-28), recurrence at the same site occurred in two patients (10.0%); amputation-free survival at 12 months was 100% and overall survival 95%. The bFGF-treated cohort carried a disproportionate burden of DM and PAD, supporting targeted early biologic intervention in this highest-risk subgroup. Within a stepwise algorithm combining objective perfusion assessment with prespecified escalation criteria, 90% of wounds healed completely without amputation, providing a practical limb-salvage framework for this vulnerable population.

Accurate estimation of ground reaction forces (GRFs) is fundamental to gait analysis and supports functional gait assessment, surgical planning, and rehabilitation. We propose a physics-learning hybrid framework that combines linked-segment dynamics with a bidirectional long short-term memory network to estimate vertical GRFs (vGRFs) from seven inertial measurement units (IMUs). The learning module produces an unconstrained allocation variable that decomposes the physics-derived net force into limb-specific vGRFs. The framework was systematically evaluated against machine learning-based baselines, including MLP, LSTM, BiLSTM, and Transformer variants, using three datasets: two self-collected datasets comprising healthy individuals and individuals with unilateral transtibial amputation, and an independent public biomechanics dataset covering multiple walking speeds, ramps, and stairs. Evaluation included intra-subject, intersubject, cross-population, and between-day evaluations, as well as cross-task transfer. The proposed framework achieved an RMSE of 0.034 BW in intra-subject evaluation and 0.065 BW in inter-subject evaluation, both below the 0.10 BW high-accuracy reference, while maintaining high coefficients of determination (R2 = 0.994 and R2 = 0.978, respectively). These findings demonstrate that integrating biomechanical structure with data-driven learning can improve vGRF estimation performance in both intra- and inter-subject settings, while showing potential for transfer across populations, measurement sessions, and locomotor tasks. The framework therefore supports further development toward quantitative gait assessment and broader clinical applications.