Notice

 
Abstract
The presentation will start with a brief introduction about the multi-disciplinary organization scheme of Dokuz Eylul University (Izmir, Turkey) on biomedical imaging research and education.
The lecture will cover challenges and recent developments in biomedical image analysis and informatics under four main steps, which are demonstrated with applications on abdomen and specifically on pre-evaluation of donors prior to liver transplantation surgery:
1. First, radiological image processing pipeline will be introduced and recent trends on using machine-learning based approaches for combining outputs of several segmentation algorithms (in order to obtain a superior result) will be summarized. In other words, the question of “how to use the outputs of different liver segmentation methods quantitatively, such that a weighted combination of them would produce more accurate results” will be addressed.
2. After a precise segmentation is obtained, the next step requires clear 3D visualization of the liver together with its adjacent tissues. Accordingly, the second topic will discuss how to use novel multi-dimensional transfer functions for generating a complete rendering of the scene without segmenting each tissue of interest. A chronological development of our studies will be summarized. Based on the outcomes, new volume enhancement approaches will be introduced.
3. Once a satisfactory 3D representation is created, it is usually stored as exported videos or snapshots. This approach not only limits the reproducibility and minor modifications at a later time, but also requires significant amount of space for storage. To overcome these shortcomings, our new DICOM compliant 3D-presentation states approach and its implementation will be discussed.
4. Finally, there will a short discussion about 3D-printing process of a segmented liver, which requires integration of individual 3D-prints of different vascular trees inside epoxy filled liver shell, and how surgeons use it to guide the operation.
 
Short Bio
Alper Selverreceived the BSc degree from Gazi University, Ankara, Turkey, in 2002, and the MSc and PhD degrees from Dokuz Eylul University (DEU), Izmir, Turkey, in 2005 and 2010, respectively, all in electrical and electronics engineering (EEE). During his graduate studies, he has studied inMedical Informatics Laboratory at FH-Aachen Abt. Juelich, Germany and Heffner Biomedical Imaging Laboratory at Columbia University, NewYork, USA. Since 2015, he has been working as an associate professor at the DEU EEE. His mainr esearch interests include the field of radiological image processing and transfer function specification for visualization, hierarchical, and mutiscale classification strategies forbiomedical applications and software development for the use of thedeveloped techniques in clinic.