Project Details

Popis obrazových dat pro podobnostní vyhledávání v medicínských databázích

Project Period: 1. 1. 2003 - 31. 12. 2003

Project Type: grant

Code: FR830/2003/G1

Agency: Fond rozvoje vysokých škol MŠMT

Program:

English title

Image Data Description for Content Based Retrieval in Medical Databases

Type

grant

Keywords

Texture analysis, image segmentation, content based retrieval.

Abstract

The Project deals with a developing of the image data description techniques in medicine, for computer tomography (CT) and magnetic resonance (MR). This description should serve for content based retrieval in medical image databases. The goal of the project is: 1. Design of the image data description technique for CT/MR images focused on texture properties of tissues and its comparison with other methods on the collection of real images. 2. Design of the technique for CT/MR image segmentation, or modification of some commonly used approach. 3. Design (or modification) of desription method (from 1) for property description of regions obtained by segmentation process. 4. Creation of the prototype implementation of a system for CT/MR image segmentation and automatical desription for the purpose of content based retrieval.

Team members

Heckel Martin, Ing. (UIFS FIT VUT) , research leader
Kršek Přemysl, doc. Ing., Ph.D. (UPGM FIT VUT) , team leader

Publications

2003

HECKEL Martin and ZENDULKA Jaroslav. Data Mining and Its Use in Texture Analysis. In: Proceedings of the CS&P'2003 Workshop. Warsaw: Warsaw University, 2003, pp. 225-234. ISBN 83-88374-71-0. Detail
HECKEL Martin. Texture Analysis via Data Mining. In: Proceedings of the 2nd Australasian Data Mining Workshop. Sydney: University of Technology Sydney, 2003, pp. 95-104. ISBN 0-9751724-1-7. Detail