Anisotropic filtering for detecting left ventricular borders in echocardiographic images

Modelling image smoothing as a nonlinear diffusion process leads to an adaptive filter method. Here it is shown that this method can successfully be applied to ultrasound images as a preprocessing step in the implementation of a left ventricular contour finder. The iterative diffusion process is implemented on a parallel transputer hardware. The contour finder uses a deformable contour model which contains smoothness and connectivity constraints. The contour detection is performed by minimizing an energy function