dscf1390 Bilge Karaçalı, PhD

Professor, Electrical and Electronics Engineering Department

Director, Biomedical Information Processing Laboratory

Izmir Institute of Technology

Urla, 35430 Izmir TURKEY

Phone : 90-232-750-6534

Fax      : 90-232-750-6599

E-mail : bilge at iyte dot edu dot tr


Continue to the lab page here.


My main interest lies in constructing quantitative measures for characterizing disease related tissue abnormality and damage using all available information sources including but not limited to bioimaging modalities, genomics, proteomics, as well as clinical history. In particular, I investigate the effects of tissue abnormality and damage associated with degenerative diseases like cancer on the appearance observed by radiological images as the macro scale and histology slides at the micro scale and to the molecular scale presentation as characterized by gene and protein sequences, gene expression profiles, and metabolite concentrations. The correlations between these appearances across macro, micro, and molecular scales reveal a much more complete picture of the cumulative degeneration and abnormality induced on the affected tissue by a disease process, allowing to predict one based on the others, and collectively depict the scope of the abnormality for diagnosis and staging, as well as prognosis and the most suitable therapeutic strategies.


My current research endeavors focus on developing effective signal processing and machine learning approaches for accurate recognition of cognitive activities from electroencephalography signals. Given that the brain accomplishes complex cognitive tasks through a network of interconnected brain regions, I believe this can be accomplished by elucidating the communication patterns specific to each cognitive task of interest as they manifest in transient inter-channel synchronization profiles. Ongoing work supported by the TUBITAK grant 117E784 awarded to my name shows promising results indicating that synchronization profiles between channel pairs and clusters offer viable recognition of motor imagery activity.


Academic work

·      Quasi-supervised learning for biomedical data analysis


·      Hierarchical motif vectors for protein alignment and functional classification


·      Quantitative Analysis of Tissue Composition in Digitized Histology Slides using Automated Texture Classification in Health and Disease Conditions


·      Vector Space Methods in the Computational Analysis of Gene and Protein Sequence Data


·      New Generation Brain-Computer Interfacing Using Transient Synchronization Between EEG Channels and Channel Clusters




·         EE101 Introduction to Electrical Engineering

·         EE333 Fundamentals of Probability and Random Processes

·         EE430 Introduction to Systems Biology

·         EE434 Biomedical Signal Processing (joint course with Zübeyir Ünlü, PhD)

·         EE436 Mathematical Foundations of Signal Processing and System Control

·         EE531 Probability and Random Processes

·         EE549 Biomedical Image Analysis

·         EE550 Computational Biology

·         EE563 Selected Topics in Electrical Engineering: Flow Cytometry Data Analysis

·         EE590 Scientific Research Methods and Ethics for Engineers


Research interests

·     Biomedical Information Processing

Automatic protein classification

Computational analysis of genetic sequences

Molecular phylogeny algorithms

Biomedical image analysis

Quantitative methods in biomedical data analysis


·     Statistical Learning

Statistical learning theory

Quasi-supervised learning

Approximation, classification, and detection

Automatic target recognition and tracking


·     Computer Vision

Surface reconstruction, shape-from-shading, photometric stereo

Multispectral and hyperspectral vision

Remote sensing


·     Mathematical Methods

Vector space methods

Independent component analysis

Multivariate ranks

Wavelets and multiscale analysis

Extreme value theory





·     Modeling/Estimating Tissue Deformations in Tumor Patients, NIH Project, R01-NS042645-04, postdoctoral fellow, 2002-2005

·     2+2+2 Workforce Education in Biotechnology and Bioinformatics, PA DCED Project, project coordinator, 2005

·     Elastic Alignment of Multimodality Medical Images using Information Theoretic Point Similarity Measures, TÜBİTAK, 108E249, project investigator, 2009-2011

·     Hierarchical Motif Vectors for Protein Alignment and Functional Classification, European Research Counsel FP7, PIRG-GA-2008-230903, project coordinator, 2009-2012





·     1992 Ranked 40th nationwide in University Entrance Examination

·     1997 Ranked 1st nationwide in Graduate Education Examination

·     2007 2nd prize, GPBA Spring Retreat Poster Competition, general category



·  Simulation of tissue atrophy (distributed here)


·  Quasi-supervised learning (distributed here)


Educational History:

·  1997 - 2002 North Carolina State University, Raleigh, North Carolina

M.S., Ph. D. in Electrical Engineering
Minor in Mathematics
Dissertation "Vector Space Methods For Surface Reconstruction From One Or More Images Acquired From The Same View With Application To Scanning Electron Microscopy" web access

·  1992 - 1997 Bilkent University, Ankara, Turkey

B.S. in Electrical Engineering


·  Yetkin and I

·  Yetkin and toys

·  Yetkin after a meal

·  Yetkin and I