Non-Radioactive Infrared Thermal Imaging and AI Applications: Across Medical Imaging Modalities for Fracture Diagnosis

Abstract

Fracture diagnosis is a critical task in emergency and orthopedic medicine, yet traditional X-ray imaging has notable limitations, including the risk of misdiagnosis in occult or subtle fractures and unnecessary radiation exposure to healthy tissues. In recent years, two technological advancements, IRT and AI, have shown considerable promise in addressing these challenges. IRT offers a non-invasive and radiation-free method for detecting localized inflammation related to fractures, while AI algorithms, particularly deep learning models, have demonstrated high accuracy in automated fracture detection from radiographic images. However, current literature review on the combined applications of these three techniques are limited. As a result, this review systematically examines the current progress in integrating IRT and AI algorithms into intelligent X-ray systems for fracture diagnosis. The principles, clinical applications, and recent innovations in these technologies are outlined, including AI-guided fracture localization, thermal image analysis, and intelligent radiation collimation. Particular attention is given to the emerging concept of multimodal systems that combine radiographs, thermal data, and AI to enhance diagnostic efficiency and reduce patient radiation burden. Furthermore, critical limitations, and future research directions in this multidisciplinary field are established and debated. This review provides an overview scope to enable the development and clinical implementation of smart imaging systems in fracture treatment.