Vessel Structure Tracking

E. Poletti, D. Fiorin, E. Grisan and A. Ruggeri. Retinal Vessel Axis Estimation through a Multi-Directional Graph Search Approach. WC 2009, IFMBE Proceedings 25/XI, pp. 137–40, Springer-Verlag, Berlin Heidelberg 2009.
D. Fiorin, E. Poletti, E. Grisan and A. Ruggeri. Fast adaptive axis-based segmentation of retinal vessels through matched filters. WC 2009, IFMBE Proceedings 25/XI, pp. 145–8, Springer-Verlag, Berlin Heidelberg 2009.
E. Grisan, A. Pesce, A. Giani, M. Foracchia, and A. Ruggeri. A new tracking system for the robust extraction of retinal vessel structure. Proc. 26th Annual International Conference of IEEE-EMBS, pp. 1620-1623, IEEE, New York, 2004

Identification and measurement of blood vessels in retinal images could allow quantitative evaluation of clinical features, which may allow early diagnosis and effective monitoring of therapies in retinopathy. A new system is proposed for the automatic extraction of the vascular structure in retinal images, based on a sparse tracking technique.

After processing pixels on a grid of rows and columns to determine a set of starting points (seeds), the tracking procedure starts. It moves along the vessel by analyzing subsequent vessel cross sections , and extracting the vessel center, calibre and direction.

When tracking stops because of a critical area, e.g. low contrast, bifurcation or crossing, a “bubble technique” module is run. It grows and analyzes circular scan lines around the critical points, allowing the exploration of the vessel structure beyond the critical areas.

After tracking the vessels, identified segments are connected by a greedy connection algorithm. Finally bifurcations and crossings are identified analyzing vessel end points with respect to the vessel structure.

Seed Point Extraction:

A number of equally-spaced (one every 10 pixels) rows and columns of the image are analyzed to search for candidate seed points. Each selected line is analyzed looking for the grey-level pattern corresponding to possible vessels: the derivative signal of a vessel profile would appear as two spikes of opposite sign close to each other, corresponding to the transitions from background to vessel (first edge) and then from vessel to background (second edge).


The tracking module is based on the analysis of consecutive cross sections of a vessel. The cross section of a vessel can be very noisy, making the recognition of the vessel boundaries difficult. To have a better signal to analyze, and using the a priori information that an ideal vessel has an almost constant greylevel pattern locally along its course, we can extract n cross sections of equal length, but with center displaced along the vessel direction.

When the procedure is not able to proceed further along a vessel segment, it invokes the Bubble analysis, trying to put new seed points beyond the area that made the tracking procedure stop. The idea of the “bubble analysis” is to look in concentric circular lines around a point. The pixels of each circular line are clustered with a fuzzy c-mean algorithm, using their graylevel values as classification feature.


The tracking strategy may result in the splitting of a single vessel into two or more segments, because of the presence of critical areas where the tracking algorithm has stopped. In order to identify those extrema candidates to be connected, we have devised a greedy algorithm.

Bifurcations and Crossings Identification:

After the connection module, identification of crossings and bifurcations may seem a trivial task, yet it requires some care. Crossings are found by identifying all pairs of vessel points that are on different sides of another vessel. Bifurcations can be divided into two types requiring different approaches: the branching and the proper bifurcations.


return to top