Getting a leg up on diabetic lower limb amputation with new AI risk prediction model
SINGAPORE, May 13, 2026 /PRNewswire/ -- Personalised risk-based monitoring may soon be the norm for patients with diabetes at high risk of lower extremity amputation (LEA). Jointly developed by Singapore General Hospital (SGH), SingHealth, and the MOH Office for Healthcare Transformation (MOHT), the LEA – Neural Network Model (LEA-Net) harnesses artificial intelligence to enable earlier interventions and reduce avoidable amputations.
LEA-Net aims to predict the risk of LEA three to five years before patients develop foot ulcers and infections, allowing care teams to intervene before irreversible tissue damage occurs. It categorises patients into low- and high-risk groups, facilitating targeted interventions and reducing wait times to see vascular specialists. In Singapore, close to nine in 10 individuals who had an LEA have diabetes, with about 85 per cent of cases preceded by foot ulcers.
LEA can be devastating, affecting patients' mobility, independence, relationships, and health. Those with diabetes or vascular disease who have undergone an LEA also face added risk of wound complications and potential loss of the other limb. The financial implications are equally significant on healthcare systems — early-stage interventions average about $25,000, whilst late-stage treatments can exceed $40,000 to $50,000 per patient.
Patients with diabetes undergo annual screenings for foot and eye health and renal function, though diabetic foot screening compliance remains the lowest of the three. Many tend to only seek medical attention when a wound or infection is already present, increasing the possibility of amputation and associated risks such as surgical complications and death.
LEA-Net was developed using anonymous data from over 830,000 SingHealth patient records, incorporating demographics, clinical conditions and medical test results. Of these, about 250,000 patient records were set aside for validation. The model demonstrated a sensitivity rate of nearly 80 per cent and specificity rate of close to 90 per cent for LEA prediction, outperforming other benchmarking models on both measures.
Model results have shown that high-risk patients can be identified during the critical window before ulceration and infection develop. It also reflects the shift towards preventive, data-driven care under Healthier SG.
LEA-Net won the People's Choice Award at the International Consortium for Health Outcomes Measurement (ICHOM) Conference 2025 in Dublin, Ireland. The team is looking to further validate the model's clinical efficacy through a pilot study involving patients from SingHealth's Diabetes Registry.
SOURCE Singapore General Hospital
Share this article