A histogram-based technique for echocardiographic image enhancement

Echocardiography is a very powerful diagnostic tool. However, the acquired images are often noisy and difficult to interpret. One well-known technique for image-enhancement is to apply predefined gray-level intensity transformations, or brightness transfer functions (BTFs). This paper presents an innovative algorithm, called ABTF (adaptive brightness transfer function), designed to optimally determine the gray-levels used for each specific echocardiographic cine-loop. The algorithm applies automatic segmentation to the initial gray-level histogram, based on curve fitting to the sum of three Gaussian functions, each of which is correlated to a different tissue type. A different procedure is applied to each segment. The ABTF algorithm has been tested on 23 cine-loops from 10 different patients, with varying degrees of echogenicity. The results are clearly superior to the original images, with a much better contrast and richer gray-levels within the cardiac muscle.