摘要:
提出一种局域基谐波频谱峰值比(LFHR)的超声图像融合方法。首先,分离超声射频信号基谐波,计算二者的局域频谱峰值比,确定融合权值;然后,加权平均基谐波成分,得到融合射频信号;最后,经过希尔伯特变换、对数压缩、插值/降采样生成B超融合图像。非均匀囊肿组织模型仿真结果表明,LFHR法获取的融合图像较原始图像的对比度、对比度噪声比、组织杂波比分别提高2.44倍、2.80倍及3.22倍,谐波成分谱能量提高83%。人体颈动脉临床实验结果进一步验证了LFHR法的有效性。综上,LFHR法获得了组织轮廓清晰内部细节丰富的B超融合图像,图像质量显著提高。
Abstract:
The ultrasound image fusion method based on the Local Fundamental and Harmonic spectral peak Ratio(LFHR) is proposed.First,the fundamental and harmonic components of the ultrasonic Radio Frequency(RF) signals are separated.The weights of both components are determined according to their spectral peak ratio.Then,the fused RF signals are produced with weighted average of the fundamental and harmonic components.Finally,the ultrasound fused images are generated by the framework of demodulation,logarithmic compression and interpolation/downsampling.Experimental results based on simulations of inhomogeneous cysts show that the contrast,contrast-to-noise ratio,and tissue-to-clutter ratio of the fused images based the proposed method are 2.44,2.80 and 3.22 times more than the original images,respectively.The harmonic energy is increased by 83%.The clinical experiments with the human carotid artery also validate the effectiveness of the proposed method.In summary,the fusion B-mode images produced by the LFHR method possess rich details,clear tissue contour,and improved quality.
下载:
