Structure-Based Evaluation Methodology for Curvilinear Structure Detection Algorithms

Curvilinear structures are useful features, particularly in medical image analysis. Typically, a pixel-wise comparison with manually specified ground truth is used for performance evaluation. In this paper we propose a novel structure-based methodology for evaluating the performance of curvilinear structure detection algorithms. We consider the two aspects of performance, namely detection rate and detection accuracy, separately. This is in contrast to their mixed handling in earlier approaches that typically produces biased impression of detection quality. The proposed performance measures provide a more informative and precise performance characterization. A series of experiments in the context of retinal vessel detection are presented to demonstrate the advantages of our approach.