A Compact Textile Monopole Antenna for Monitoring the Healing of Bone Fractures Using Un-Supervised Machine Learning Algorithm

In this paper, a novel approach to monitoring the healing process of bone fractures using a compact textile monopole antenna integrated into a wearable device. A planar wearable textile ultra-wideband monopole (TUM) antenna is proposed as a sensing element. The antenna is fabricated on the jeans fabric with a thickness of 1 mm, an overall size of
$18\times19$
mm2, and operates in an ultra-wideband frequency range (3.22-10.9 GHz). The human arm tibia model is developed in a full-wave simulator in order to estimate the radiation exposure caused by the antenna at near and far field distances. At both field distances, the average specific absorption rate (SAR) values of the model are below the FCC limit. The proposed antenna is experimentally tested on a donor bovine tibia. The proposed system utilizes Unsupervised machine learning (ML) techniques to analyze the received signals from the antenna and provide real-time feedback on the progression of bone fracture healing. The compact and flexible nature of the textile antenna allows for comfortable and unobtrusive wear, making it suitable for long-term monitoring. The integration of Unsupervised ML techniques enables automated analysis of the received signals, eliminating the need for manual interpretation. This work aims to improve the monitoring and assessment of bone fracture healing, leading to more effective treatment strategies and faster recovery times. The experimental results and comparison study with previous literature show that the proposed TUM antenna has several advantages and is suitable for diagnosing bone fractures.