Petia Radeva's Home Page

Petia Radeva
Universitat Autònoma de Barcelona
Dept. Informàtica, Computer Vision Center
Edifici O
08193 Bellaterra (Barcelona)
Catalunya - Spain
Tel. 34-93-581 1862 - Fax. 34-93-581 1670
http://www.cvc.uab.es/~petia/
e-mail:petia<at>cvc<dot>uab<dot>es

"There's only one way of live and that's your own!"

---

Publications

---

Current Projects:

Project “Tissue and Dynamic Characterization of IVUS” Intravascular ultrasound is a medical imaging technique that provides a unique view of the blood vessels. Morphology, composition, location and behaviour of the different kind of tissues are shown as a cross-view section image. This project is concerned with the analysis of different phenomena involved in the vessels and the characterization of the different component tissues based on this imaging technique. Ranging from radio-frequency analysis and computation, image feature extraction and 3D reconstruction and motion and dynamic behavioural study, the MLV group is devoted to provide answers in all of these fields using advanced machine learning and computer vision techniques.

Project “Intestine Motility Analysis by Advanced Machine Learning” Intestinal motility assessment with video capsule endoscopy arises as a novel and challenging clinical fieldwork. This technique is based on the analysis of the patterns of intestinal contractions obtained by labelling all the motility events present in a video provided by a capsule with a wireless micro-camera, which is ingested by the patient. In this project, we develop advanced machine learning techniques to automatically detect the intestinal contractions in video capsule endoscopy, driving a useful but not feasible clinical routine into a feasible clinical procedure. Our proposal is based on a sequential design involving the analysis of textural, colour and blob features with powerful classifiers such as SVM, Adaboost, Cascade of classifiers, ECOC, etc. Special attention is paid to the study of the problem as an imbalance data set, as well as on the performance measures achieved by our system both in a qualitative and a quantitative way by means of several performance metrics.

---

Past Projects:

Project “Tagged MRI Analysis by Computer Vision Techniques”: MRI Tagging is a unique imaging technique that prints a grid over the tissue which moves according to the underlying movement of the myocardium revealing the true deformation of the cardiac muscle. In this project, we developed novel techniques to track grids based on spectral analysis and mutual image information combined with deformable B-solids in order to obtain the 3D local and global deformation of the myocardium.

CorkInspect is an IST EU project within the EUTIST-IMV cluster. Corkinspect is a Trial project intending to support, as a major objective, the early customisation and validation of a Computer Vision solution for cork stopper production control. The main discriminant feature of CorkInspect is its innovative development based on its way of building inspection and control rules. CorkInspect establishes its control rules through an iterative procedure based on statistical control features and boundaries tolerances from selected samples. That brings great flexibility and, in special, when we work with natural products, this flexibility to establish control features becomes a must.

CorkInspect

PharmIdent

PharmaInspect

---

Teaching Activities

Lógica Computacional

Estructura de Datos

 Course for Graduate students on Model-Based Analysis - Deformable Models

---

Open research positions (job offers)

The Computer Vision Center (http://www.cvc.uab.es/) at the Universitat Autònoma de Barcelona invites applications for 1 Postdoc and 1 PhD student in the area of Object Recognition and Medical Imaging.

The approved candidates will be involved in a Computer Vision project to develop novel advanced methods based on object and pattern recognition (multiple classifiers, mutual information, semi-supervised techniques, video analysis, etc.) to analyse endoscopic video images of intestines in order to detect dynamic patterns of intestine motility. The project counts with very advanced image acquisition technology with promising results and great interest from the clinical community. The project is being developed in a narrow collaboration with one of the most well-known clinical experts in the field coming from the Hospital "Vall d'Hebron", Barcelona Spain, as well as the support of the industrial partner possessing a very novel video acquisition system.

The research profile of the candidates should cover as much as possible:
- High motivation, scientific spirit and initiative
- Teamwork and communications skills
- Ph.D. for the postdoc position and Bachelor degree for the predoc position in computer vision and/or image processing.
- Knowledge of computer science engineering fundamentals (data structures, algorithm design and analysis, operating systems, and software engineering)
- Programming experience in Matlab (appreciated)
- Written and verbal communication skills in English.

The position is for one year, extensible to another one. The financial conditions depend on the scientific expertise of the candidates.

Applicants should submit:
- Curriculum Vitae (including photograph, personal data, grades, publications, etc.),
- names and email addresses of 3 references(the referees might be contacted),
- links to their thesis and/or publications.

Positions will remain open until filled. The applications should be sent directly to the following e-mail address: petia<at>cvc<dot>uab<dot>es (to substitute at by @ and dot by .).