Nakagami imaging with small windows

Nakagami imaging is an ultrasonic tissue characterization technique which has shown promise for differentiating different breast masses. Two parameters describing the shape and the scale of the distribution of the RF-data envelope are computed on sliding windows. The choice of the size of the windows is critical as both oversized and undersized windows can lead to an unstable estimation of the parameters. In this work, we propose to use Gamma kernel density estimation to compute a smooth estimation of the pdf within small windows. The Nakagami parameters are then computed on the estimated probability density function (pdf). Experiments on homogeneous materials show that images computed with this technique are smoother, though bias occurs. Application on clinical ultrasound images of breast and liver tumors demonstrate the ability of the proposed method to produce perceptually plausible parametric images.