MeDEISA : Medical Database for the Evaluation of Image and Signal Processing Algorithms
MeDEISA is a medical image and signal database, proposed by the « Engineering in Medicine and Biology » (EMB) French Chapter of the IEEE, Section France. It can be used as a tool to perform objective evaluations of image and signal processing algorithms (such as coding, denoising, segmentation,…). Furthermore, it can be used for performance method comparisons as well as medical data analysis purposes.
MeDEISA is upgradeable, in the sense that it will be frequently updated, using data provided by various clinical centres or hospitals (Please contact us if you want to participate in the evolution of MeDEISA).
Our objective consists in gathering the various acquisition modalities (preferably recent ones) of Magnetic Resonance Imaging, X-rays, Echography, Scintigraphy, with various human organs, various pathologies and various degrees of subtleties.
Each image, each spatial, temporal or spacio-temporal sequence is proposed in various formats (DICOM, PNG useful for compression purpose...), jointly with a descriptive text file that specifies the features of each medical image/signal.
The access to MeDEISA for downloading the images/signals is totally free of charge. We made it available for scientist communities, obviously for non commercial purposes.
The French EMB chapter highly encourages the researchers, engineers and students to evaluate their algorithms on the images/signals of this database.
The French EMB chapter also encourages the partnerships. Any proposal will be welcomed (Please contact us).
Contact:
Christine CAVARO-MENARD : christine.menard@univ-angers.fr
Université d’Angers, Équipe Signal-Image-LISA-EA 4014, Angers, FRANCE.
Amine NAIT-ALI : naitali@univ-paris12.fr
Université Paris 12, LISSI, EA 3956, Créteil, FRANCE.
Partnerships:
- La Société Française de Radiologie.
- Centre Hospitalier Universitaire d’Angers.
- Centre Hospitalier Universitaire de Créteil.