Abstract
Purpose
The goal is to automatically detect anomalous vascular cross-sections to attract the radiologist’s attention to possible lesions and thus reduce the time spent to analyze the image volume.
Materials and methods
We assume that both lesions and calcifications can be considered as local outliers compared to a normal cross-section. Our approach uses an intensity metric within a machine learning scheme to differentiate normal and abnormal cross-sections. It is formulated as a Density Level Detection problem and solved using a Support Vector Machine (DLD-SVM). The method has been evaluated on 42 synthetic phantoms and on 9 coronary CT data sets annotated by 2 experts.
Results
The specificity of the method was 97.57% on synthetic data, and 86.01% on real data, while its sensitivity was 82.19 and 81.23%, respectively. The agreement with the observers, measured by the kappa coefficient, was substantial (κ = 0.72). After the learning stage, which is performed off-line, the average processing time was within 10 s per artery.
Conclusions
To our knowledge, this is the first attempt to use the DLD-SVM approach to detect vascular abnormalities. Good specificity, sensitivity and agreement with experts, as well as a short processing time, show that our method can facilitate medical diagnosis and reduce evaluation time by attracting the reader’s attention to suspect regions.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Adame IM, van der Geest RJ, Wasserman B, Mohamed M, Reiber JHC, Lelieveldt BPF (2004) Automatic plaque characterization and vessel wall segmentation in magnetic resonance images of atherosclerotic carotid arteries. In: Fitzpatrick JM, Sonka M (eds) Proceedings of SPIE Med. Imaging 2004: image processing, vol 5370. San Diego, pp 265–273
Barnett V, Lewis T (1994) Outliers in Statistical Data, 3rd edn. John Wiley & Sons, West Sussex
Bellman R (1961) Adaptive control processes. Princeton University Press, New Jersey
Chang CC, Lin CJ (2001) LIBSVM: a library for support vector machines. Software available at http://www.csie.ntu.edu.tw/~cjlin/libsvm
Cohen J (1960) A coefficient of agreement for nominal scales. Educ Psychol Meas 20(1): 37–46. doi:10.1177/001316446002000104
Dehmeshki J, Ye X, Wang F, Lin XY, Abaei M, Siddique MM, Qanadli SD (2004) An accurate and reproducible scheme for quantification of coronary artery calcification in ct scans. In: Proceedings of 26th annual international conference IEEE EMBS. San Francisco, California, USA, pp 1918–1921
Fleiss JL (1971) Measuring nominal scale agreement among many raters. Psychol Bull 76(5): 378–382
Frangi AF, Niessen WJ, Vincken KL, Viergever MA (1998) Multiscale vessel enhancement filtering. In: Wells WM, Colchester A, Delp SL (eds) Medical image computing and computer-assisted intervention–MICCAI’98, Lecture Notes in Computer Science, vol 1496. Springer Verlag, Berlin, Germany, pp 130–137
Hernández Hoyos M, Orkisz M, Douek PC, Magnin IE (2005) Assessment of carotid artery stenoses in 3D contrast-enhanced magnetic resonance angiography, based on improved generation of the centerline. Mach Graph Vis 14(4): 349–378
Hernández Hoyos M, Serfaty JM, Maghiar A, Mansard C, Orkisz M, Magnin IE, Douek PC (2006) Evaluation of semi-automatic arterial stenosis quantification. Int J Comput Assist Radiol Surg 1(3): 167–175
Hush D, Kelly P, Scovel C, Steinwart I (2005) Provably fast algorithms for anomaly detection. Technical Report LA-UR-05-4367, Los Alamos National Laboratory
Išgum I, Rutten A, Prokop M, van Ginneken B (2007) Detection of coronary calcifications from computed tomography scans for automated risk assessment of coronary artery disease. Med Phys 34(4): 1450–1461
Kang DG, Suh DC, Ra JB (2009) Three-dimensional blood vessel quantification via centerline deformation. IEEE Trans Med Imaging 28(3): 405–414
Krissian K, Maladain G, Ayache N (2000) Model based detection of tubular structures in 3D images. Comput Vis Image Unders 80: 130–171
Landis JR, Koch GG (1977) The measurement of observer agreement for categorical data. Biometrics 33(1): 159–174
Lekadir K, Yang GZ (2006) Carotid artery segmentation using an outlier immune 3D active shape models framework. In: Larsen R, Nielsen M, Sporring J (eds) MICCAI 2006, LNCS 4190. Springer, Berlin, Heidelberg, Copenhagen, pp 620–627
Lesage D, Angelini ED, Bloch I, Funka-Lea G (2009) A review of 3D vessel lumen segmentation techniques: Models, features and extraction schemes. Med Image Anal 13(6): 819–845
Renard F, Yang Y (2008) Image analysis for detection of coronary artery soft plaques in MDCT images. In: 2008 IEEE International Symposium Biomed. Imaging: from nano to macro (ISBI 2008). Paris, France, pp 25–28
Rinck D, Krüger S, Reimann A, Scheuering M (2006) Shape-based segmentation and visualization techniques for evaluation of atherosclerotic plaques in coronary artery disease. In: Cleary KR, Galloway Jr RL (eds) Proceedings of SPIE Med. Imaging 2006: Visualization, image-guided procedures and display, vol 6141, pp 124–132
Salvado O (2006) Characterization of atherosclerosis with magnetic resonance imaging, challenges and validation. Ph.D. thesis, Case Western Reserve University
Schaap M, Metz C, van Walsum T, van der Giessen AG, Weustink AC, Mollet NR, Bauer C, Bogunovic H, Castro C, Deng X, Dikici E, O’Donnell T, Frenay M, Friman O, Hernández Hoyos M, Kitslaar PH, Krissian K, Kühnel C, Luengo-Oroz MA, Orkisz M, Smedby Ö, Styner M, Szymczak A, Tek H, Wang C, Warfield SK, Zambal S, Zhang Y, Krestin GP, Niessen WJ (2009) Standardized evaluation methodology and reference database for evaluating coronary artery centerline extraction algorithms. Med Image Anal 13(5): 701–714
Schölkopf B, Platt JC, Shawe-Taylor J, Smola AJ, Williamson RC (2001) Estimating the support of a high-dimensional distribution. Neural Comput 13: 1443–1471
Steinwart I, Hush D, Scovel C (2005) A classification framework for anomaly detection. J Mach Learn Res 6: 211–232
Steinwart I, Hush D, Scovel C (2005) Density level detection is classification. Neural Inf Process Syst 17: 1337–1344
Vukadinovic D, van Walsum T, Rozie S, de Weert T, Manniesing R, van der Lugt A, Niessen WJ (2009) Carotid artery segmentation and plaque quantification in CTA. In: Proceedings of 6th IEEE International Symposium Biomed. Images (ISBI’09). Boston, USA, pp 835–838
Wong WCK, Chung ACS (2006) Augmented vessels for quantitative analysis of vascular abnormalities and endovascular treatment planning. IEEE Trans. Med. Imaging 25(6): 665–684
World Health Organization: The top ten causes of death—fact sheet N310 (2008)
Wörz S, Rohr K (2007) Segmentation and quantification of human vessels usind a 3D cylindrical intensity model. IEEE Trans Image Process 16(8): 1994–2004
Zhao F, Zhang H, Wahle A, Thomas MT, Stolpen AH, Scholz TD, Sonka M (2009) Congenital aortic disease: 4D magnetic resonance segmentation and quantitative analysis. Med Image Anal 13(3): 483–493
Author information
Authors and Affiliations
Corresponding author
Additional information
This work has been supported by the ECOS-Nord Committee, project C07M04, by the Région Rhône-Alpes (France) via the Simed project of the ISLE research cluster, and by the project CIFI-Uniandes No. 54. M.A. Zuluaga’s PhD project is supported by a Colciencias grant.
Rights and permissions
About this article
Cite this article
Zuluaga, M.A., Magnin, I.E., Hernández Hoyos, M. et al. Automatic detection of abnormal vascular cross-sections based on density level detection and support vector machines. Int J CARS 6, 163–174 (2011). https://doi.org/10.1007/s11548-010-0494-8
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11548-010-0494-8