Automated Pressure Ulcer Lesion Diagnosis for Telemedicine Systems

Abstract

The timely diagnosis and treatment of pressure ulcers is a critical task and constitutes a challenge in patient rehabilitation. In this work we investigate how a tool for automated pressure ulcer stage classification can be integrated into an asynchronous telemedicine system aiming to increase efficiency and monitoring capabilities for large volumes of patient data. The deployment requirements, the internal architecture as well as the employed techniques are outlined. Furthermore, the initial processing results are provided to demonstrate the feasibility of automated classification of pressure ulcer regions in various grades.

Enhanced Assessment of the Wound-Healing Process by Accurate Multiview Tissue Classification

“…Several classification tools have also been addressed and supervised ones have been shown to surpass the others [6], [34]. In many studies, assessment of the tissue type is not achieved (only the grade or the surface of the wound is provided) [20], [26], [43] or only partially.…”

“…Pioneer experiments in this area consisted simply in uploading images to a web site where a physician could view the data at his convenience [30]. In more recent studies, image processing has been added but it provides only ulcer stage grading [26]. The quantitative assessment of chronic wounds still relies on visual inspection and manual techniques to describe the shape of the wound (perimeter, surface, depth, etc.)…”

Robust tissue classification for reproducible wound assessment in telemedicine environments

“…17,23 Several classification tools have also been addressed and supervised tools have been shown to surpass the others. 24,25 In many studies, 2,26,27 assessment of the tissue type is not achieved ͑only the grade or the surface of the wound is provided͒ or is only partially accomplished.…”

“…1 In more recent studies, image processing has been added, but it provides only ulcer stage grading. 2 The quantitative assessment of chronic wounds still relies 3,4 on visual inspection and manual techniques, to describe the shape of the wound ͑perimeter, surface, depth, etc.͒ and the biological nature of the skin tissues ͑percent-age of each class, wound severity stage, burn degree, etc.͒. Wound dimensions and shape are currently measured with an ordinary ruler, or sometimes through sketches on crossruled sheets, serum injection, or alginate moldings.…”

Fuzzy spectral clustering for automated delineation of chronic wound region using digital images

“…Most of the algorithms related to wound assessment involves high computational cost and are not real time application which limits their uses into smart phones and tablets. The feasibility of these diagnostic tools for telemedicine system have been investigated by many researchers [24,25].…”