| Accurate arterial centerline extraction is essential for comprehensive visualization in lower extremity CT angiography in patients with Peripheral Arterial Occlusive Disease (PAOD). Time-consuming manual tracking is needed when automated methods fail to track centerlines through severely diseased and occluded vessels. A method containing two algorithms is presented in this thesis to tackle the problem. The first algorithm, Partial Vector Space Projection (PVSP), is used for the estimation of missing data given a database of similar datasets. The algorithm performs Principal Component Analysis (PCA) on a database of centerlines to obtain a set of orthonormal basis functions defined in a scaled and oriented frame of reference, and assumes that any curve not in the database can be represented as a linear combination of these basis functions. We demonstrate its use in restoring the centerlines through simulated occlusions of femoropopliteal arteries, derived from CT angiography data. We evaluated the algorithm using a database of centerlines derived from 30 normal femoropopliteal arteries by deleting portions of each centerline for several Occlusion Lengths (OL: 10mm, 25mm, 50mm, 75mm, 100mm, 125mm, 150 mm, 175mm, and 200mm) and compared it to a correlation-based linear Minimum Mean Squared Error (MMSE) method. The results were fairly accurate even for large occlusion lengths and are clinically useful. The results were also consistently better than those using the MMSE method. Multivariate regression analysis found that OL and the root-mean-square error in the 2cm proximal and distal to the occlusion accounted for most of the error.; The PVSP algorithm produces encouraging results, but it is not sufficiently accurate for higher occlusion lengths (> 100mm). Results show that when a point on the centerline mid-way in the occlusion is accurately known, and is incorporated in the PVSP algorithm, the MD error was reduced by an average of 43%. Such an intermediate point can be found using calcified plaque that may form on the inner wall of the occluded artery. We introduce a new algorithm, Intermediate Point Detection (IPD), which identifies calcified plaque in image space to find the most useful point within the occlusion to improve the estimate from PVSP. In this algorithm, candidates for calcified plaque are automatically identified on axial CT slices in a restricted region around the estimate obtained from PVSP. A modified Canny edge detector identifies the edge of the calcified plaque and a convex polygon fit is used to find the edge of the calcification bordering the wall of the vessel. The Hough transform for circles estimates the center of the vessel on the slice, which serves as an intermediate point candidate. Each candidate is characterized by three scores based on the radius of the circle fitted and the relative position of the candidate within the occluded segment. Three polynomial weighting functions (types A, B, and C) are constructed, each one representing one possibility to define a net score representing the potential benefit of using this candidate for improving the centerline. This approach was tested in 44 femoropopliteal artery occlusions of lengths up to 398mm, in 30 patients with peripheral arterial occlusive disease. Centerlines were tracked manually by 4 experts, twice each, their mean serving as the reference standard. All occlusions were first interpolated with PVSP using a database of femoropopliteal arterial shapes obtained from 59 subjects, as a leave-on-out validation. Occlusions longer than 80mm (N=20) were then processed with the IPD algorithm, provided calcifications were found (N=14). MD error was used as the error metric. The IPD algorithm with type A weighting function reduced the average error from 2.76mm to 1.76mm. The error was less than the clinically desirable 3mm limit (smallest radius of the femoropopliteal artery) in 13 of 14 occlusions. The results were within 1mm of the highest human reade... |
