Time Deformable Segmentation Model Based on the Active Contour Driven by Gaussian Energy Distribution: Extraction and Modeling of Early Articular Cartilage Pathological Interuptions

摘要

In the clinical orthopaedics, the articular cartilage monitoring is an important task having especially preventive effect. The magnetic resonance (MR) is commonly used clinical standard allowing for the effective differentiation of articular cartilage from surrounding tissues (bones, soft tissues). Nevertheless, the early pathological interruptions are often badly recognizable from the native MR records. This fact significantly influences clinical diagnosis. We have carried out the analysis of the segmentation method based on the active contour with the aim of autonomous modelling articular cartilage and indication of the early cartilage interruptions. The active contour model represents time deformable model adopting the articular cartilage geometrical features with respect to cartilage interruptions. Model of the articular cartilage reflects area of the physiological cartilage in the form of binary segmentation while the active contour model is terminated in the spot of the early pathological sign. Therefore, this time deformable model has ambitions to be used as a feedback to subjective physician's opinion because the model clearly differentiates the physiological cartilage structure from the early cartilage loss.