Fusion of thermal and three-dimensional data for chronic wound monitoring

Abstract

Chronic wounds are a serious health problem because they do not follow a normal healing process and take a long time to heal. Monitoring the evolution of the wound is vital to evaluate the current treatment and determine the need for a different approach. Traditionally, this assessment is based on manual measurements and visual observation, which can be subjective and non-quantitative, leading to uncertainty in patient follow-up. This thesis proposes to leverage color and thermal imaging from cameras embedded in a smartphone to provide a non-invasive tool for wound assessment. Standard and low-cost devices ensure its applicability and scalability in clinical settings. We propose a methodology to create realistic and accurate 3D thermal models from a limited set of images captured with portable devices and using freehand shooting. In addition, the proposed methodology was tested in several cases of wounds, and we show a proof of concept of the measurements obtained with the thermal 3D model for wound evaluation in a clinical study, in a medical center in Lima, Peru.