Is thermography a viable solution for detecting pressure injuries in dark skin patients?

TL;DR

A new thermal and optical imaging dataset of 35 participants focused on darker skin tones where temperature differences are induced through cooling and cupping protocols and the preliminary results suggest that thermography-based CNN is robust to data collection protocols for all skin tones.

Abstract

Pressure injury (PI) detection is challenging, especially in dark skin tones, due to the unreliability of visual inspection. Thermography has been suggested as a viable alternative as temperature differences in the skin can indicate impending tissue damage. Although deep learning models have demonstrated considerable promise toward reliably detecting PI, the existing work fails to evaluate the performance on darker skin tones and varying data collection protocols. In this paper, we introduce a new thermal and optical imaging dataset of 35 participants focused on darker skin tones where temperature differences are induced through cooling and cupping protocols. We vary the image collection process to include different cameras, lighting, patient pose, and camera distance. We compare the performance of a small convolutional neural network (CNN) trained on either the thermal or the optical images on all skin tones. Our preliminary results suggest that thermography-based CNN is robust to data collection protocols for all skin tones.

Figures (7)

• Figure 1: Example of control images from our dataset with the 4 different skin tones.
• Figure 2: Bar plot of misclassified optical images based on image collection protocol.
• Figure 3: Progression of erythema in thermal image from $0-7$ minutes post-cupping. PC denotes Post-Cupping, square represents the cooling spot, circle represents the erythema spot.
• Figure 4: Workflow for classifying cooling or erythema images.
• Figure 5: Confusion matrix of the three image modalities in the cooling classification task.