Linxuan Zhou, Shuangyi Cheng, Kailiang Xu
Resolution quantification has gained considerable attention for ultrasound localization microscopy (ULM). However, due to microvasculature heterogeneity, there is a clear requirement of characterizing the distribution of local and anisotropic resolution. Aiming for that, the rolling Fourier ring correlation (rFRC) and the rolling full-width at half-maximum (rFWHM) methods are introduced to characterize the resolution distribution on pixel scale of the entire ULM image. In addition, the rolling Fourier line correlation (rFLC) method is introduced for quantifying the anisotropic resolution on pixel level. Potential applications of the proposed resolution distribution are further explored, including assessing the performance of the ULM reconstruction algorithms, determining optimal super-resolution pixel size, and guiding data acquisition time.