GMM4MEDICAL

Adaptive Gaussian Mixture Models for Continuous Representation of Digital Medical Images

Coordinatore	RHEINISCH-WESTFAELISCHE TECHNISCHE HOCHSCHULE AACHEN    more  Organization address address: Templergraben 55 city: AACHEN postcode: 52062 contact info Titolo: Prof. Nome: Ernst Cognome: Schmachtenberg Email: send email Telefono: +49 241 8090490 Fax: +49 241 8092490
Nazionalità Coordinatore	Germany [DE]
Totale costo	30˙000 €
EC contributo	30˙000 €
Programma	FP7-PEOPLE Specific programme "People" implementing the Seventh Framework Programme of the European Community for research, technological development and demonstration activities (2007 to 2013)
Code Call	FP7-PEOPLE-2009-RG
Funding Scheme	MC-ERG
Anno di inizio	2010
Periodo (anno-mese-giorno)	2010-06-01   -   2012-05-31

 Partecipanti

#	participant	country	role	EC contrib. [€]
1	RHEINISCH-WESTFAELISCHE TECHNISCHE HOCHSCHULE AACHEN  Organization address address: Templergraben 55 city: AACHEN postcode: 52062 contact info Titolo: Prof. Nome: Ernst Cognome: Schmachtenberg Email: send email Telefono: +49 241 8090490 Fax: +49 241 8092490	DE (AACHEN)	coordinator	30˙000.00

Mappa

 Word cloud

Esplora la "nuvola delle parole (Word Cloud) per avere un'idea di massima del progetto.

 Obiettivo del progetto (Objective)

'In tomographic medical imaging, images are not acquired directly but sample data of statistical nature is measured from the patient placed in the field of view. From the acquired data, a volumetric image is then reconstructed by computational methods. Since the data acquisition pattern does not take into account the underlying image representation, reconstruction artifacts are likely to occur, especially when images are represented by uniform grids of voxels. As a consequence, images contain visible noise artifacts while the resolution is often insufficient in regions that would be supported by higher statistical information. Those regions are the focus of attention for image assessment and improving image resolution locally could provide a huge benefit for better detectability. Alternatives to the classical representation of images by grids of pixels and voxels exist but image modeling is not yet a very active field of research. Fortunately, the combination of modern developments in statistical estimation methods, approximation properties of polynomial B-spline basis functions and efficient hierarchical space partitioning data structures provide both theoretical justifications and efficient computational methods for the generation of high-quality adaptive image models from limited input data. The aim of this research project is to unveil a new way to represent continuous digital images in general. The paradigm of continuous image representation is totally new for medical imaging and contrasts with established discrete image models based on histograms. With such a sparse and continuous model, the image space is not limited by sharp boundaries and the number of image elements, hence the resolution, can be adapted locally as a function of the amount of input information available for image reconstruction. The techniques developed in this project will have a strong impact since they can be transferred to many other stochastic reconstruction scenarios.'