During angioplasty, a guide wire (GW) is routinely placed in the coronary artery. Balloon inflation during angioplasty causes transient occlusion of the coronary artery and regional dysfunction. Thus, it is of major importance to monitor myocardial function, which may be impaired during this period. Since the GW moves with the coronary arteries, information regarding myocardial function can potentially be extracted from the GW motion. An algorithm is suggested which is a step toward such monitoring. The algorithm presented is a semiautomatic procedure for identifying and tracking the GW using specific characteristics of the GW. This algorithm is based on working in limited active windows. A preprocessing stage which enhances the GW by the use of a modified Laplacian filter or a modified Marr-Hildreth filter is introduced. The second stage of the algorithm is the tracking of the GW, which is based on fitting a second-degree polynomial to the GW using the Hough transform in each window. To further improve the results further modifications of the basic algorithms that were taken. A single set of parameters, which enabled good tracking for a large number of images taken during angioplasty, was fitted to the final algorithm.