Infection identification by traditional clinical signs and symptoms detection methods is prone to variation and error among clinicians. However, hyperspectral imaging could allow clinicians to detect and intervene with infections earlier. In a multi-center prospective study of 66 outpatient wound care patients, hyperspectral imaging with Swift’s Ray 1 collected visible light, thermography and bacterial fluorescence images. The wounds were also evaluated and screened using the International Wound Infection Institute checklist for clinical signs and symptoms of infection. Principal component analysis was conducted on the images to identify wounds as infected, inflamed or non-infected.

The trained model with features from Ray 1 images, accurately predicted all three wound classes (infected, inflamed and non-infected) with a 74% accuracy rate. When differentiating between infected vs. non-infected wounds, the model correctly identified 100% of infected wounds and 91% of non-infected wounds. 

Read more: https://www.frontiersin.org/articles/10.3389/fmed.2023.1165281/full