Automatic vessel segmentation in MR angiography

Abstract Funding Acknowledgements Type of funding sources: None. Background Vessel-to-background contrast (VBC), the ratio signal intensities (SI) within a vessel segment compared to surrounding tissue, is an essential objective parameter for assessing image quality in contrast-enhanced MR angiography (CE-MRA) peripheral arteries. (1–3) Manual evaluation by measurement SIs and area labour-intensive tedious. (2) An algorithm automatic segmentation lumen background reduces workload significantly. It has already been successfully applied CE-MRA arterial disease (PAD). (3) However, whether this equal manual not yet known. Purpose To compare VBC patients with PAD. Methods A patient suspected underwent first-pass subtraction MRA on 3T scanner. Region-of-interests (ROIs) were drawn predefined segments (Figure 1), manually inside outside additionally comprehensive ROI under investigation centre. The segmented into identifying seed point then using flood fill algorithm. remaining branches removed order approximate true signal. SI define formula - divided standard deviation (SD) both methods. Results 43 evaluated. example given Figure 2 A) showing axial view right thigh superficial femoral artery centre (green), outlining (red) (blue). (B1) (B2) used evaluate manuall values. Automated C1) shows C2) full extent as identified flood-fill algorithm, C3) resulting mask. was calculated from C4 after local maximas shown C5). mean 4171 (standard deviation, SD 954) 5014 (SD 830) (p = 0.001) 601 542) 493 229) ((p 0.231). 10.3 5.9) 12.2 5.6) 0.139). Conclusion Automatic allows assessment tends be overestimated about 20% Higher numbers are required further clarify applicability evaluation.