Research Seminars

Visual Computing Research Seminar

Research Laboratories

Intelligent Multimedia Processing (IMP)

Medical Informatics (MedIX)

The main two medical imaging projects our group is focusing on are lung nodule interpretation and soft tissue segmentation.

Project 1: Lung Nodule Interpretation
Studies show that lung cancer detection and interpretation improves when multiple radiologists read the CT scans. CAD (Computer Aided Diagnosis) systems can act as a second opinion for the radiologist. Our current research goals include; determining the relationship between physical nodule characteristics and their annotations. Understanding how radiologists rate nodules and how they interact with CAD. The goal is to create Content-Based and Semantic-Based Image Retrieval Systems (CBIR and SBIR) for helping in the education of future radiologist. Click here for a handout or contact Ekarin Varutbangkul (realekarin@gmail.com) / William Horsthemke (horst@ameritech.net) for more information.

Project 2: Soft Tissue Segmentation
Proper segmentation of soft tissues in medical images is necessary for a variety of medical tasks; for example doctors need to know an organ’s volume and location before performing a transplant operation. Proper segmentation is also an important early step to developing a CAD (Computer Aided Diagnosis” program. Organs share similar intensity values and are located in close proximity to each other, fooling many segmentation algorithms. Texture solves these problems because while organs have similar intensity values, their texture is often unique. Our goals for this project include; creating a single filter for each organ and to then combine these filters; determining if 3D texture algorithms collect more information than 2D texture algorithms, and developing a probabilistic algorithm to classify organs. Click here for a handout or contact Ruchaneewan Susomboon (rsusombo@students.depaul.edu) for more information.
 

back to top

Our research approaches a patent database and its citations with theoretical models applied from statistical analysis and physics.  Using the clustering techniques of complex networks, we search for patterns of relationships between patents of different categories.  The research will allow legal specialists to identify the authority and importance of particular patents and measure their difference across various categories (i.e., the extent of difference between biotech and software categories).  Using the citations of patents to identify “truly authoritative” patents across different years and categories, our research measures the value of innovations and their impact in various fields of industry.  Intellectual property specialists from DePaul’s College of Law contribute the analysis, guidance, and instruction necessary to successfully identify the measures. 

back to top

Bioinformatics

Argonne National Laboratory (ANL) through its Biochip Technology Center funded $50,000 to the CTI’s Intelligent Multimedia Processing (IMP) Lab to support Image Processing and Data Analysis research work within the context of the Department of Homeland Security (DHS) microbial forensics programs. The IMP Lab will work towards the development of statistically-based experimental designs, microarray image analysis and decision tools for the analysis of genotyping and single nucleotide polymorphism microarrays.

Microarray Toolbox Download, click here

back to top

NSF MedIX: Medical Informatics Experiences for Undergraduates

The NSF REU site is in the area of Medical Informatics and will address issues related to the workflow driven acquisition of image data in an electronic medical record environment, and the subsequent post-processing of the image information to improve the accuracy and efficiency of the consumers of this information, notably radiologists. In particular, segmentation, texture analysis, classification and annotation of computed tomography (CT) images using computer vision and data mining techniques will be emphasized.

The MedIX REU site will run for three summers, with eight student participants spending 10 weeks each summer doing research both at DePaul University, Intelligent Multimedia Processing Laboratory and Northwestern University, Feinberg School of Medicine, Imaging Informatics Laboratory.  Student participants will be paid a stipend of $4000 for 10 weeks, $400 meal allowance, plus travel support to present their summer work at a conference.  For more information, click here.

back to top

Intelligent Multimedia Processing (IMP) Laboratory hosts the undergraduate and graduate students doing Visual Computing research at CTI; there are around 10 students performing research in the lab every year.  The IMP lab facilities include six brand new high-end workstations boasting Intel P4 HT processors at 3.2 GHz and having large amounts of memory (1GB ~ 2GB).   These facilities were provided by CTI to encourage the efforts made by the VC group to promote undergraduate and interdisciplinary research.  The undergraduate activities that have been already established in the IMP lab include: 1) an annual orientation and research projects’ overview, 2) weekly meetings, 3) written reports, and 4) presentations to the Visual Computing research group seminar and the UPE undergraduate research seminar. 

Medical Informatics Processing (MedIX) Laboratory

The MedIX lab houses eight workstations providing workspace for eight full-time students.  The MedIX workstations have the latest P4 3.2 GHz HT processors 1GB of DDR each.   The lab was funded by CTI to support the NSF REU site on Medical Informatics at CTI.

back to top