Our Mission

Development of the new image processing and machine learning methods for medical image analysis.

Faculty of Electrical Engineering, Computer Science and Information Technology Osijek

----------------------------------------------------------------------------------------
All our researchers are employed @ Faculty of Electrical Engineering, Computer Science and Information Technology Osijek

Croatian science foundation


----------------------------------------------------------------------------------------
This work has been supported in part by Croatian Science
Foundation under the project UIP-2017-05-4968.

WP1 - Segmentation and analysis of heart chambers and aorta anatomy

Automatic segmentation of heart chambers directly depends on the localization of these chambers within a CT or an MRI image. A geometry chambers analysis (volume of a heart muscle, minute volume, ejection fraction) directly depends on the appropriate segmentation.

More Information

WP2 - Segmentation and analysis of atrium and left atrial appendage

Segmentation and a left atrial appendage analysis depend on the precise determination of atrial segmentation after which anatomic structures raising from an atrium (primarily pulmonary veins and a left atrial appendage) are detected.

More Information

WP3 - Segmentation and analysis of other relevant coronary arteries and epicardial fat quantification

Automatic segmentation of heart chambers directly depends on the localization of these chambers within a CT or an MRI image. A geometry chambers analysis (volume of a heart muscle, minute volume, ejection fraction) directly depends on the appropriate segmentation.

More Information

WP4 - Simulation of a blood flow through heart parts

A method simulating a blood flow through heart structures will be developed in the fourth work package. The method will focus on the simulation of a blood flow through an aorta and a left atrial appendage.

More Information

WP5 - Isolation and the entire heart visualization

The previously developed methods will be used to implement isolation and heart visualization methods in the fifth work package. These methods are of the utmost importance to doctors in order to determine heart anatomy as well as the location and geometry of coronary arteries. The heart isolation methods stem from previous segmentation of heart structures using the adjustment of predefined heart mesh and machine learning.

Learn More

Learn With Us

Check out our blog full of tutorials regarding medical image processing, machine and deep learning!
en_USEnglish (United States)
hrCroatian en_USEnglish (United States)