Data-driven analysis of structure borne sound for the detection of hip-stem implant loosening
One major reason for required revision of total hip replacements is loosening of the hip-stem. This is often detected late due to low sensitivity and specificity of current diagnostic methods. As alternative it as has been suggested to analyze the structure borne sound in the hip-stem generated by a forced vibration excitation. For this purpose, an existing hip-stem implant was modified to house a mechanical oscillator which can be excited by an extracorporeal magnetic coil. The excitation system is directly located in the hip-stem implant thus allowing for in vivo monitoring of implant loosening. In previous studies using artificial bones is was shown that loosening states can be classified successfully by sensing the structure borne sound with vibration sensors. In order to investigate the inter-individual variability resulting from anatomy, experiments with four human femur bones from body donations have been conducted. The aim of this contribution is to use the experimental data in order to develop robust classifiers detecting the loosening of hip-stem using deep learning techniques. The development of the classifiers as well as their in-depth evaluation is discussed.