Classification of Pressure Ulcer Images with Logistic Regression

Abstract

Pressure ulcers are wounds caused by prolonged lying in bedridden patients. This has become an important health problem in many countries. Correct diagnosis of pressure ulcers is very important for an effective treatment method. The characteristics of these wounds are effective in terms of seeing the healing. Interventional methods of obtaining information in the diagnosis of pressure ulcers are painful for patients. In addition, these methods can increase the risk of infection. Therefore, imaging systems such as nonsurgical wound tracking techniques allow accurate analysis of the features of the wound without contact with it. The aim of this study is to prevent wound formation or to make a positive contribution to the treatment processes by using machine learning techniques in image analysis for the classification of pressure ulcers. In this study, 142 wound images were analyzed by Logistic Regression and Artificial Neural Networks methods. Features such as wound color and size in these images were separated by image processing and the stage of the wound was determined from the images. The 6 stages of pressure ulcers are referenced for classification.