Researchers at the German Cancer Research Center (DKFZ ) and the Department of Anaesthesiology at Heidelberg University Hospital (UKHD) have developed an innovative approach for the early and precise detection of sepsis. This serious complication, which is the third most common cause of death in Germany, often occurs after oncological operations. The new method uses hyperspectral imaging in combination with artificial intelligence (AI) to diagnose sepsis quickly, non-invasively and without taking blood samples.

Hyperspectral imaging captures light reflections from the skin on the palms and fingers, which make changes in microcirculation and in the concentration of water, blood and oxygen visible. This data is converted into clinically relevant information using neural networks. In a study with over 480 critically ill patients in the intensive care unit of the UKHD, the AI-based method achieved a prediction quality of 0.80, which increased to 0.94 when vital parameters were added. The prediction of 30-day mortality was also highly accurate.

Compared to established clinical scores and biomarkers such as C-reactive protein or procalcitonin, the technology enables faster and more reliable screening, especially in time-critical situations. The method is mobile, cost-effective and does not burden patients. It shows potential for use in emergency rooms, in the ambulance service or for the early detection of complications after operations. The researchers are planning further studies to test its clinical effectiveness, including in other patient groups such as children. The work was funded by the ERC Consolidator Grant, the Hidss4Health project and the Andreas Hoeft Grant from the European Society of Anaesthesiology and Intensive Care, among others.

Original Paper:

AI-powered skin spectral imaging enables instant sepsis diagnosis and outcome prediction in critically ill patients | Science Advances

Read also:

Emergency medicine: Rescue services recognize sepsis far too rarely – MedLabPortal

Half of all patients with sepsis die within two years – MedLabPortal

Molecular pathogen diagnostics in sepsis: faster to targeted therapy – MedLabPortal

Editorial office: X-Press Journalistenbüro GbR

Gender note. The personal designations used in this text always refer equally to female, male and diverse persons. Double/triple references and gendered designations are avoided for the sake of better readability ected.