Imaging Biomarkers

Our research is focused on the development of new segmentation and classification techniques that allow the diagnosis and follow up of cardiac and vascular pathologies. Using our in-house segmentation algorithms, we seek to combine the information provided by multiple image modalities to improve the segmentation and posterior analysis of different structures of interest. 

According to the World Health Organization, cardiovascular diseases account for around 30% of deaths around the world. The use of medical image computing in clinical routine has shown a tremendous potential to reduce the death toll by allowing early diagnosis and treatment of the different pathologies.


Cardiac Image Analysis

We have developed a multi-atlas segmentation propagation (MASP) approach for whole heart segmentation that has been proven to be robust to different image modalities. At the moment, we are using this method to build a framework that guides the segmentation of the left atrium in ultrasound images by using a priori information from MR and CT, and in computer-assisted diagnosis of congenital heart diseases (Fig. 1).


figure 2 


Fig. 1 CAD pipeline for congenital heart disease diagnosis based on whole heart segmentation. For further details refer to Zuluaga et al, ISBI 2014.

Vascular Image Analysis 

vesselsSC2 web

In a similar fashion, we try to exploit the information from multiple images to improve the quality of the segmentation of vascular structures. We have been applying such techniques to different applications:

  • Cerebrovascular tree extraction for epilepsy surgery planning.
  • Plaque image analysis and correlation using ex-vivo MR and histology.
  • OCT segmentation.

Our main collaborator in the area of vascular image analysis is the Institute of Neurology at UCL. 


Our Collaborators

Silvia Schievano - UCL Institute of Cardiovascular Science
Andrew Taylor - Great Ormond Street Hospital
Jim Duncan - Yale University
Rolf Jager - Institue of Neurology (IoN)
Kawal Rhode - King's College London



  1. A. Atehortúa, M.A. Zuluaga, S. Ourselin, E. Romero. Automatic segmentation of 4D cardiac MR images for extraction of ventricular chambers using a spatio-temporal approach. In: SPIE Medical Imaging (In press)
  2. B. Biffi, M.A. Zuluaga, S. Ourselin, A.M. Taylor, S. Schievano. Papillary Muscle Segmentation from a Multi-Atlas Database: A Feasibility Study. STACOM 2015, LNCS 9534, 1-10 (2016)
  3. M.A. Zuluaga*, N. Burgos*, A.F. Mendelson, A.M. Taylor, S. Ourselin. Voxelwise Atlas Rating for Computer assisted Diagnosis: Application to Congenital Heart Diseases of the Great Arteries. Medical Image Analysis 26(1) 185-194 (2015)
  4. M.A. Zuluaga, R. Rodionov, M. Nowell, S. Achhala, G. Zombori, A.F. Mendelson, M.J. Cardoso, A. Miserocchi, A.W. McEvoy, J.S. Duncan, S. Ourselin. Stability, Structure and Scale: improvements in Multi-modal Vessel Extraction for SEEG Trajectory Planning. IJCARS 10(8) - Special Edition on MICCAI 2014, 1227-1237 (2015)
  5. M.A. Zuluaga*, N. Burgos*, A.M. Taylor, S. Ourselin. Multi-atlas Synthesis for Computer Assisted Diagnosis: Application to Cardiovascular Disease. ISBI 2015. New York, USA. pp. 290-293 (2015)
  6. C. Tobón-Gómez, A.J, Geers, J. Peters, J. Weese, M.A. Zuluaga, K. Rhode, et al. Benchmark for Algorithms Segmenting the Left Atrium from 3D CT and MRI Datasets.  IEEE Transactions on Medical Imaging 34 (7), 1460-1473 (2015)
  7. C. Tobon-Gomez, M.A. Zuluaga, H. Chubb, S.E. Williams, C. Butakoff, R. Karim, O. Camara, S. Ourselin, K. Rhode. Standardised unfold map of the left atrium: regional definition for multimodal imaging. Journal of Cardiovascular Magnetic Resonance 17 (Supp 1) P41 (2015)
  8. C. Petitjean, M.A. Zuluaga, et al. Right Ventricle Segmentation From Cardiac MRI: A Collation Study. Medical Image Analysis19(1),  187-202 (2015)
  9. M.A. Zuluaga, R. Rodionov, M. Nowell, S. Achhala, G. Zombori, M.J. Cardoso, A. Miserocchi, A.W. McEvoy, J.S. Duncan, S. Ourselin. SEEG Trajectory Planning: Combining Stability, Structure and Scale in Vessel Extraction. In: Medical Image Computing and Computer-Assisted Interventions - MICCAI.  Part II, LNCS 8674, pp. 651-658 (2014) 
  10. MA. Zuluaga, A. Mendelson, M.J. Cardoso, A.M. Taylor, S. Ourselin. Multi-Atlas Based Pathological Stratification of D-TGA Congenital Heart Disease. In: 11th International Symposium on Biomedical Imaging - ISBI 2014. Beijing, China. pp. 109-112 (2014)
  11. M.A. Zuluaga, M.J. Cardoso, M. Modat, S. Ourselin. Multi-Atlas Propagation Whole Heart Segmentation from MRI and CTA Using a Local Normalised Correlation Coefficient Criterion. In: Functional Imaging and Modeling of the Heart - FIMH. LNCS 7945, 174-181 (2013)
  12. M.A. Zuluaga, M.J. Cardoso, S. Ourselin. Automatic Right Ventricle Segmentation Using Multi-Label Fusion in Cardiac MRI. Workshop on RV Segmentation Challenge. MICCAI (2012)
  13. X. Zhuang, K.S. Rhode, R.S. Razavi, D.J. Hawkes, S. Ourselin. A Registration-based Propagation Framework for Automatic Whole Heart Segmentation of Cardiac MRI. IEEE Transactions on Medical Imaging 29(9), 1612-1625 (2010) 
  14. X. Zhuang, K.S. Rhode, S. Arridge, R.S. Razavi, D.Hill, D.J. Hawkes, S. Ourselin. An Atlas-based Segmentation Propagation Framework Using Locally Affine Registration - Application to Automatic Whole Heart Segmentation. In: Medical Image Computing and Computer-Assisted Interventions - MICCAI 2008, 425-433 (2008)