<![CDATA[Ultrasound imaging is one of the most widely used non-destructive testingmethods. The transducer emits pulses that travel through the imaged samplesand are reflected by echo-forming impedance. The resulting ultrasonic signalsusually contain noise. Most of the traditional noise reduction algorithmsrequire high skills and prior knowledge of noise distribution, which has acrucial impact on their performances. As a result, these methods generallyyield a loss of information, significantly influencing the final data and deeplylimiting both sensitivity and resolution of imaging devices in medical andindustrial applications. In the present study, a denoising method based on anattention-gated convolutional autoencoder is proposed to fill this gap. Toevaluate its performance, the suggested protocol is compared to widely usedmethods such as butterworth filtering (BF), discrete wavelet transforms(DWT), principal component analysis (PCA), and convolutional autoencoder(CAE) methods. Results proved that better denoising can be achievedespecially when the original signal-to-noise ratio (SNR) is very low and thesound waves’ traces are distorted by noise. Moreover, the initial SNR wasimproved by up to 30 dB and the resulting Pearson correlation coefficient wasmaintained over 99% even for ultrasonic signals with poor initial SNR.]]>
Copyrights © 2023
