在小动物PET成像系统中,脉冲峰值精度决定着能量分辨率和位置分辨率,对成像质量极为重要。我们通过精确仿真评估了LMS自适应滤波方法峰值估计的性能,结果表明相对传统直接取最大值方法,在窄脉冲及较大噪声时可以给出很高的精度;并且由于简单通用,该方法在数字峰值估计上有很高的应用价值。
The accuracy of pulse height estimation is important in small animal PET, because it determinates the energy resolution and spacial resolution. In the paper, it evaluates the performance of LMS adaptive filtering based pulse height estimation through accurate simulation. The result demonstrates that this method shows immune to noise and can give a more accurate height value. Also, the simplicity and versatility make it useful in pulse height estimation.