1. INTRODUCTION
Improving ultrasound resolution is an ongoing issue and many studies have been performed for this purpose; see e.g. [1], [2], [3], [4], [5], [6]. We report here a prospective study in which our aim was to contribute to improvement of the resolution ultrasound images using multidimensional parametric autoregressive (AR) modeling. An N dimensional complex number AR model represents the components of the complex number signal at locations as a linear combination of the complex number components and an additive noise , where , and is a set of neighbors excluding (0, 0, …, 0). We recently proposed a least-square-based technique to estimate multidimensional AR parameters which was able to estimate both orders and parameters of a general N-dimensional model, in the presence of white noise [7]. In order to improve the resolution of ultrasound images, in this study we extended this method by using an instrumental variable-type algorithm which is able to deal with the colored noise (due to artifacts) present in ultrasound images. This paper is structured as follows: first, we provide a short description of ultrasound imaging, then we present the proposed multidimensional technique, and finally we give an illustration of the improvement in resolution on ultrasound images.