Wavelet decomposition processing method : Cortical bone thickness measurement using ultrasound sensor

Wavelet analysis has been successfully applied in many applications such as image processing, communication and signal processing. Moreover, the wavelet theory covers more significant application. It is because the wavelet theory can be applied to the continuous and the discrete-time cases. In particular, the wavelet transform focuses on the analysis of non-stationary signals. When compared with classical Short-Time Fourier Transform and Fourier Transform analysis, both processing methods are unable to process the non-stationary signal. While wavelet analysis provides an alternative solution for this problem. Ultrasound signal is categorized as a non-stationary signal. Many techniques have been used for processing the ultrasound signal. The use of ultrasound in medical and civil area keeps increasing due to low cost, and accurate techniques. Therefore, this study focuses on the analysis of the multiple echoes of ultrasound in the cortical bone using wavelet decomposition processing method. Then, the appeared echo at both surfaces of bone was determined using a peak detection algorithm. Bone thickness was calculated based on measured flight time in the bone. The experimental results were compared with manual measurement using vernier caliper for validation process. The results show that the wavelet decomposition processing method provided small errors of cortical bone thickness measurement, which was 0.152 mm (4.104%), while standard deviation was 0.102 mm. This measurement values showed that the accuracy of bone thickness was almost 96.000% for this method. Furthermore, the experimental results were correlated with caliper values so that correlation coefficient was 0.9797.