1. Extraction of Vessels from X-Ray Angiograms Titus Rosu Prof. Dr. Rupert Lasser Andreas Keil

2. Extraction of Vessels from X-Ray Angiograms2 Introduction Applications Initial step for feature-based algorithms –Intra-operative guidance –Reconstruction of coronary arteries Objective of the IDP Segmentation of coronary arteries Implementing different algorithms in the ITK framework Testing and numerical evaluation between the algorithms

3. Extraction of Vessels from X-Ray Angiograms3 Introduction Data (Rotational) angiography sequences from stationary C-arms Contrasted coronary arteries No radial distortion Pixel spacing is 0.3 x 0.3mm or 0.6 x 0.6mm Problems Projection images  Overlay of vessels and other structures Varying contrast Decreasing vessel width

4. Extraction of Vessels from X-Ray Angiograms4 Methods Improving the images with image processing Algorithms (thresholding, normalizing, cropping...) Segmentation Algorithms Multiscale vessel enhancement filtering – (Frangi, 1998) Multiscale detection of curvilinear structures in 2-D and 3-D image data – (Koller, 1995) Vessel segmentation using a shape driven flow – (Nain, 2004) Segmentation Assumption of a linear structure of the vessels Eigen value analysis of the image intensities of every pixel Analyzing with different scales (scale space)

5. Extraction of Vessels from X-Ray Angiograms5 Methods Nain Region based flow deforms the curve of interest Using level set techniques to evolve the active contour Determine the shape of a contour with a local ball filter (values between 0-1) small near-circle evolution User input: Max. vessel width Nain, Yezzi, and Turk. Vessel segmentation using a shape driven flow. MICCAI, vol. 3216 of LNCS, pp. 51-59. Springer, 2004

6. Extraction of Vessels from X-Ray Angiograms6 Methods Frangi One of the standard papers on Hessian-based vessel filtering Basis for many other papers w.r.t –Usage of the Hessian –Multiscale analysis for vessels Frangi, Niessen, Vincken, and Viergever. Multiscale vessel enhancement filtering. MICCAI, vol. 1496 of LNCS, pp. 130-137. Springer, 1998

7. Extraction of Vessels from X-Ray Angiograms7 Methods Koller Detect curvilinear structures of arbitrary shape Using the 2 nd derivation of the Gauß-Function to resolve the edges left and right of the vessel-profile Non linear algorithm => using positive min-function Mulitscale analysis User input min. and max. of the vessels width Koller, Gerig, Székely, and Dettwiler. Multiscale detection of curvilinear structures in 2-D and 3-D image data. ICCV, pp. 864-869, 1995

8. Extraction of Vessels from X-Ray Angiograms8 Post processing and Segmentation Frangi image post processing: Double threshold => Segmentation

9. Extraction of Vessels from X-Ray Angiograms9 Post processing and Segmentation Koller image post processing: centerline detection => => double threshold => => connected line det.=> segmentation C. BLONDEL, G. MALANDAIN, R. VAILLANT,, and N. AYACHE, Reconstruction of Coronary Arteries From a Single Rotational X-Ray Projection Sequence, IEEE TRANSACTIONS ON MEDICAL IMAGING, VOL. 25 (2006), pp. 653-663.

10. Extraction of Vessels from X-Ray Angiograms10 Post processing and Segmentation Koller image post processing: double threshold => segmentation

11. Extraction of Vessels from X-Ray Angiograms11 Evaluation Quantitative comparison of results Manually segmented data as “ground truth” Same comparison conditions by post processing the result images with different filters

12. Extraction of Vessels from X-Ray Angiograms12 Evaluation 3 images from different data sets for comparison Comparison: double threshold Frangi vs. double threshold Koller vs. ground truth double threshold Frangi vs. post processed centerline Koller vs. ground truth Numerical pixel wise comparison:

13. Extraction of Vessels from X-Ray Angiograms13 Evaluation min. connected lines of 5 pixels Koller image min. connected lines of 20 pixels Koller image Double thres. Koller imageDouble thres. Frangi image Data setResult FN Result FP Result TP Result FN Result FP Result TP Result FN Result FP Result TP Result FN Result FP Result TP 71011C0.570.360.430.600.270.400.530.420.470.550.410.45 70822E0.590.210.410.610.160.390.550.200.450.520.170.48 70822B0.590.370.410.630.260.370.560.330.440.540.300.46 best vs. worst

14. Extraction of Vessels from X-Ray Angiograms14 Evaluation Merged double thres. Frangi, min. con. lines of 5 pixels Koller image Merged double thres. Frangi, min. con. lines of 20 pixels Koller image Merged double thres. Frangi, double thres. Koller image Data setResult FN Result FP Result TP Result FN Result FP Result TP Result FN Result FP Result TP 71011C0.440.600.560.460.54 0.430.640.57 70822E0.450.290.550.460.240.540.440.260.56 70822B0.440.510.560.460.420.540.440.460.56 Top: Frangi vs. Koller Bottom: Merged Frangi, Koller vs. manually seg. images Overlapped pixels are segmented white, x = min(Frangi(x), Koller(x)) Merged Frangi, Koller images produce better TP/FN but increase FPs best vs. worst

15. Extraction of Vessels from X-Ray Angiograms15 Evaluation From top to bottom: Frangi vs. man. seg. images Min. connected component lines of 5 pixels Koller vs. man. seg. images Min. connected component lines of 20 pixels Koller vs. man. seg. images Double threshold Koller vs. man. seg. images Overlapped pixels are segmented white, x = max(Frangi(x), Koller(x))

16. Extraction of Vessels from X-Ray Angiograms16 Evaluation From top to bottom: Merged double threshold Frangi, min. connected component lines of 5 pixels Koller vs. man. seg. images Merged double threshold Frangi, min. connected component lines of 20 pixels Koller vs. man. seg. images Merged double threshold Frangi, Koller vs. man. seg. images Overlapped pixels are segmented white, x = min(Frangi(x), Koller(x)) Merged Frangi, Koller images produce better TP/FN but increase FPs

17. Extraction of Vessels from X-Ray Angiograms17 Conclusion Lesser noisy connected pixel areas Detects better the vessel width Frangi Detects better smaller vessels Segments the bones thinner Koller Both delivers good results, better results maybe with optimized constants, post processing algorithms

18. Extraction of Vessels from X-Ray Angiograms18 Conclusion Lesser noisy connected pixel areas Detects better the vessel width FrangiKoller Detects better smaller vessels Segments the bones thinner

19. Extraction of Vessels from X-Ray Angiograms19 Conclusion Lesser noisy connected pixel areas Detects better the vessel width FrangiKoller Detects better smaller vessels Segments the bones thinner

20. Extraction of Vessels from X-Ray Angiograms20 Conclusion Lesser noisy connected pixel areas Detects better the vessel width FrangiKoller Detects better smaller vessels Segments the bones thinner