Purpose:  High rates of amputation and mortality are associated with diabetes-related foot disease (DRFD) and chronic limb-threatening ischemia (CLTI) worldwide. In Australia, DRFD alone contributes to an estimated 2500 deaths and 6300 lower limb amputations annually. AI methods offer powerful new approaches to multi-modal predictive modelling but require large and well-characterised datasets. We have created a registry of DRFD and CLTI patients in SA in order to enable epidemiological modelling, health services research, and the development of AI-powered predictive models to identify patients at high risk of amputation. Methods: Our statewide retrospective cohort study is being constructed using routinely collected de-identified clinical data comprising electronic medical records, medical imaging, and bespoke surgical datasets. The dataset includes patient demographics, comorbidities, wound characteristics, perfusion assessments, treatment details, and radiological imaging.  Results:  Our project will provide invaluable insights into the epidemiology of CLTI and DRFD in South Australia. By training AI models on a large and diverse dataset, the project will deliver tangible clinical benefits by developing robust predictive models applicable to a wide range of patients, able to identify key predictive factors of amputation risk and mortality. The registry will also establish a framework for translational medical AI research. Conclusion:  By detailing project methodology, discussing the logistical challenges, and presenting epidemiological data on the historical cohort of patients with CLTI and DRFD in South Australia, the feasibility and benefits of this approach can be discussed, as well as the opportunity for further national expansion of the registry. Ultimately, this project aims to improve risk stratification, guide early interventions, reduce amputation rates, and optimise healthcare resources, thereby enhancing patient outcomes in DRFD and CLTI.