We present in this paper the application of deep convolutional neural networks (CNNs), which is a state-of-the-art artificial intelligence (AI) approach in machine learning, for automated time-independent prediction of burn depth. Color images of four types of burn depth injured in first few days, including normal skin and background, acquired by a TiVi camera were trained and tested with four pretrained deep CNNs: VGG-16, GoogleNet, ResNet-50, and ResNet-101. In the end, the best 10-fold cross-validation results obtained from ResNet-101 with an average, minimum, and maximum accuracy are 81.66, 72.06, and 88.06%, respectively; and the average accuracy, sensitivity, and specificity for the four different types of burn depth are 90.54, 74.35, and 94.25%, respectively. The accuracy was compared with the clinical diagnosis obtained after the wound had healed. Hence, application of AI is very promising for prediction of burn depth and, therefore, can be a useful tool to help in guiding clinical decision and initial treatment of burn wounds.

The perfusion in scalds between 72-96h after injury, as measured using LSCI, is highly predictive of healing outcome in scalds when measured. The predictive value can be further improved by incorporating an early perfusion measurement within 24h after injury.

BackgroundLaser Speckle Contrast Imaging (LSCI) is a non-invasive and fast technique for measuring microvascular blood flow that recently has found clinical use for burn assessment and evaluation of flaps. Tissue motion caused by for example breathing or patient movements may however affect the measurements in these clinical applications, as may distance between the camera and the skin and tissue curvature. Therefore, the aims of this study were to investigate the effect of frame rate, number of frames/image, movement of the tissue, measuring distance and tissue curvature on the measured perfusion.MethodsMethyl nicotinate-induced vasodilation in the forearm skin was measured using LSCI during controlled motion at different speeds, using different combinations of frame rate and number of frames/image, and at varying camera angles and distances. Experiments were made on healthy volunteers and on a cloth soaked in a colloidal suspension of polystyrene microspheres.ResultsMeasured perfusion increased with tissue motion speed. The relation was independent of the absolute perfusion in the skin and of frame rate and number of frames/image. The measured perfusion decreased with increasing angles (16% at 60°, p = 0.01). Measured perfusion did not vary significantly between measurement distances from 15 to 40 cm (p = 0.77, %CV 0.9%).ConclusionTissue motion increases and measurement angles beyond 45° decrease the measured perfusion in LSCI. These findings have to be taken into account when LSCI is used to assess moving or curved tissue surfaces, which is common in clinical applications.

Background: Microvascular perfusion changes in scalds in children during the first weeks