Automating Wound Assessment: Convolutional Neural Network Interface for WIFI and SINBAD Classification Systems

This study aimed to develop a mobile application incorporating the SINBAD and WIFI classification systems for diabetic foot ulcers, targeting healthcare professionals in wound care clinics to facilitate accurate and efficient diagnosis and management. A total of ۲۳۳ pictures of patients with diabetic foot ulcers, labeled by physicians and nurses from an outpatient wound clinic, were included in the analysis. We implemented a multitask, multi-output Convolutional Neural Network (CNN) model, utilizing the MobileNetV۳ Small pretrained model as its base to assess the components of the SINBAD and WIFI scoring systems. The CNN model demonstrated high performance in classifying wound-related conditions, suggesting its potential for integration into clinical diagnostics to enhance decision-making in patient care. Additionally, the user interface of the application is designed for the easy classification of diabetic foot ulcers using the trained model, providing immediate feedback and improving accessibility for healthcare professionals, thus serving as a valuable tool for real-time assessment and clinical decision-making.

کلیدواژه ها:

Wound Assessment ، Convolutional Neural Network ، SINBAD Classification System ، WIFI Classification System ، Diabetic Foot Ulcers ، Healthcare Professionals ، Mobile Application ، Clinical Diagnostics