Matching layer design of a 2–2 piezo-composite ultrasonic transducer for biomedical imaging

The performance of the piezo-composite transducer is greatly influenced by the matching layers. Based on finite element analysis and artificial intelligence, an optimization design method is proposed to optimize the matching layers of a 2–2 piezo-composite transducer for biomedical imaging. The neural networks are trained by the finite element analysis data to build up the mapping relationship between the thickness of matching layers and the performance. The optimization criteria are established based on the performance (centre frequency, bandwidth, and peak-to-peak voltage) and minimizing the material consumption. The thicknesses of matching layers are optimized by using a particle swarm optimization algorithm. The optimized thicknesses of both matching layers are 49 μm. The optimized centre frequency, bandwidth, and peak-to-peak voltage are 9.1 MHz, 72.5%, and 2.26 V, which can nearly achieve the designed targets (9.5 MHz, 75%, and 2.20 V). According to the optimized thicknesses of matching layers, the fabricated 2–2 piezo-composite transducer exhibits a centre frequency of 9.8 MHz, a bandwidth of 79.6%, and a peak-to-peak voltage of 2.11 V verifies the effectiveness and availability of the proposed method. Then, the piezo-composite transducer is hard-pressed for ultrasonic imaging. The press-focused piezo-composite transducer has an insertion loss of -19.1 dB at 10 MHz, a lateral resolution of 125 μm at 4 mm, and an axial resolution of 132 μm. The good resolution was confirmed by scanning the pig eyeball, which suggests that the 2–2 piezo-composite transducer with optimized matching layers has great potential in biomedical imaging.