Research

Quantification of lung ventilation in free breathing MRI

Development of an automated method to quantify lung ventilation in free breathing 3D+t MRI scans. The MATLAB scripts rely on image registration and the Fourier transform to perform the ventilation analysis on each voxel.

[To be published]

Image biomarker discovery and quantification

Development of a fully automated method (C++ and MATLAB) that, given a thoracic chest 3D CT, segments, measures, and pairs airway branches with the accompanying artery. It then quantifies airway wall thickening and bronchiectasis by measuring the wall-artery ratio, airway-artery ratio, and airway tapering. Automatically obtained biomarkers aggregated for each patient can separate diseased from controls as well as manual annotation do. In addition, biomarkers from each airway branch can be used to colour-code the visualisation of 3D rendered airway segmentations to highlight differences in airway structure.

Automatic airway-artery pairing on lung CT to quantify airway wall thickness and bronchiectasis
A. Perez-Rovira, W. Kuo, J. Petersen, H. Tiddens, M. de Bruijne.
Vol. 43, Pp. 5736-5744 (10). Medical physics. 2016.

Quantification of diaphragm mechanics in 3D+t MRI

Use of image registration and lung surface extraction to quantify lung kinematics during breathing. This allows to quantify and analyse the separate contribution of diaphragm and thoracic muscles.

Automatic Quantification of Diaphragm Mechanics in Pompe Disease using Dynamic 3D MRI
K. Mogalle, A. Perez-Rovira, P. Ciet, S.C.A. Wens, et al.
11 (7), e0158912. doi:10.1371/journal.pone.0158912. PLOS ONE, 2016.

Lumen calibre measurement for stenosis detection and visualisation

groba Development of GroBa, a lumen calibre measurement technique based on growing balloons. GroBa presents the advantages of cross-sectional based methods, as it is able to cope with irregular, non-tubular vessel structures, such as stenosis or aneurysms, but at the same time it is able to obtain precise calibre measurements even when the estimated centrelines are not accurate. The technique was applied to full-body angiographic MRI data, where diameter measurements were used to enhance visualisations.

GroBa: Growing balloons for calibre measurement on stenotic lumens
A. Perez-Rovira, E. Trucco, J. Weir-Mccall, G. Houston,
25th IEEE International Symposium on Computer-Based Medical Systems (CBMS). Rome (Italy), 2012.

Registration of fluorescein angiogram sequences of the retina

rerbee Development of RERBEE (Robust Efficient Registration via Bifurcations and Elongated Elements), a feature-based registration algorithm able to correct local deformations in high-resolution ultra-wide field-of-view (UWFV) fluorescein angiogram (FA) sequences of the retina. The algorithm is able to cope with peripheral blurring, severe occlusions, presence of retinal pathologies and the change of image content due to the perfusion of the fluorescein dye in time. GPU was used to increase the performance of the most computationally expensive parts of the algorithm by a factor of over ×1300.

RERBEE: Robust Efficient Registration via Bifurcations and Elongated Elements
applied to retinal fluorescein angiogram sequences

A. Perez-Rovira, R. Cabido, E. Trucco, S. McKenna and J.P. Hubschman.
Vol 31(1), Pp. 140-150. doi: 10.1109/TMI.2011.2167517. IEEE Transactions on Medical Imaging. 2012.

Retinal vessel segmentation

Vessel segmentation on ultra-wide field-of-view fluorescein angiogram sequences of the retina using steerable filters and automatic thresholding. A penalization stage is introduced to cope with regions showing high vessel response in the filtered image. Code and data are publicly available at the VAMPIRE website. The public data consists of 8 frames from 2 different sequences, including original frame, manual segmentation, and the automatic segmentation results.

Improving vessel segmentation in ultra-wide field-of-view retinal fluorescein angiograms
A. Perez-Rovira, K. Zutis, J.P. Hubschman and E. Trucco.
Pp 2614-2617. 33rd IEEE Engineering in Medicine and Biology Conference. Boston (USA). 2011.

VAMPIRE

vampiregui VAMPIRE (Vessel Assessment and Measurement Platform for Images of the REtina) is a software tool for efficient, semi-automatic quantification of retinal vessel properties on large collections of fundus camera images that I started developing during my PhD that has grown to a 10-strong person group funded by a £1,000,000 grant. [BBC and The Guardian]. The system provides automatic detection of retinal landmarks (optic disc, vasculature), and quantifies key parameters used frequently in investigative studies: vessel width, branching. and tortuosity. More information is available at the VAMPIRE website.

VAMPIRE: Vessel Assessment and Measurement Platform for Images of the REtina
A. Perez-Rovira, T. MacGillivray, E. Trucco, K.S. Chin, K. Zutis, et al.
Pp 3391-3394. 33rd IEEE Engineering in Medicine and Biology Conference. Boston (USA). 2011.

Optic disc detection

od-located Robust optic disc location by combining the response of multiple weak detectors (main arcades, macula and optic disc) on retinal fundus images.

Robust optic disc location via combination of weak detectors
A. Perez-Rovira, E. Trucco.
Pp. 3542-3545. 30th IEEE Engineering in Medicine and Biology Conference. Vancouver (Canada). 2008.

Video mosaicing, vehicle tracking, and ground plane estimation

In collaboration with Tecnobit, I developed MATLAB tools for automated video mosaicing and vehicle tracking. In addition, as part of my MSc thesis, I developed and algorithm that relied on structure from motion methods to detect the vehicle yaw rotation and estimate the ground plane orientation it was laying on.


Segmentation, registration, and manipulation of 3D point-based LADAR scenes

As part of a project for the CIDA (Centro de investigacion y desarrollo de la armada) I coded a Windows application (C++ with MFC and openGL) that automatically stiched multiple 3D scenes, segmented vehicles, estimated and tessellated the ground, and allowed for intaractive manipulation of the 3D LADAR data.

Group dynamics for videogame characters

For my BSc thesis, I wrote a Windows application (C++ with MFC and openGL) that simulated group dynamics for videogame characters navigating through a walled scene.