Brief description:

The overall goal of this Category I proposal is to develop a software tool which allows quantitative assessment of structural and functional changes during the progression of medically-refractory pediatric epilepsy in order to improve the localization of epileptic foci for surgical treatment. Towards this goal, our Institution has pioneered the use of positron emission tomography (PET) in children with epilepsy and has made important contributions [1-5]. These achievements have not only improved patient care in the State of Michigan, but have resulted in our Institution emerging as a major national and international referral center for the localization and surgical treatment of intractable epilepsy in children. However, despite improvements in seizure focus localization and improved surgical outcome, children with extratemporal lobe epilepsy still do not achieve a 90% seizure-free outcome as is possible with temporal lobe epilepsy surgery. Surgical failures may be partly accounted for by two major causes: (a) the site of seizure onset depicted by intracranial EEG is often not coincident with the location of abnormal PET areas, but is typically adjacent to it, and (b) there are often additional areas of abnormal PET in cortex remote from, but in areas that are synaptically connected to, the primary seizure focus. Therefore, major reasons for surgical failure may involve inadequate coverage by intracranial electrodes of the primary focus, and the presence of such ‘secondary' epileptic foci. The present proposal seeks to integrate the new method of diffusion tensor imaging (DTI) into the clinical assessment of secondary epileptic foci through the development of a software tool which is able to quantify the connection strength between distinct cortical areas and to spatially relate this connection strength to both functional PET imaging as well as to electrophysiological measures in pediatric epilepsy. The development of this software tool is important and significant as it may lead to new insights into the formation, identification, connectivity and behavior of secondary epileptic foci, which are the major cause for failure in pediatric epilepsy surgery.

Participants:

Otto Muzik, PhD.(PI)

Jing Hua, PhD.(Co-PI)

Shiyong Lu, PhD.(Co-PI)

Collaborative Company:

Lakeshore Engineering Services, Inc.

Kevin Parikh

Prosanto Hazra

Sam Shah

Consultant:

Susumu Mori, PhD. School of Medicine, Johns Hopkins University

Advisory board:

Farshad Fotouhi, PhD.

Csaba Juhasz, MD. PhD.

Resources:

ET/MR Imaging Laboratory. All data acquisition and testing of the software tool will be performed at the PET/MR Imaging laboratory. Clinical PET and MR imaging is performed daily in order to provide the physicians with imaging clues with respect to the location and size of primary and secondary epileptic foci. The availability of this data is paramount for a successful development of the proposed software tool. The PET Center is a 6000 square feet facility located at Children's Hospital of Michigan and part of the Wayne State University Medical School. The $5M PET/cyclotron/radiochemistry complex houses a state-of-art SIEMENS EXACT HR high-resolution whole body PET scanner and a CTI 11 MeV cyclotron. The chemistry laboratory is equipped with the necessary equipment needed to produce and develop PET radiopharmaceuticals for clinical use. Directly adjacent to the PET Center is the MRI facility which includes a 1.5 T GE Horizon MR system (GE Medical Systems, Milwauukee, WI) equipped with a 33 mT/m gradient subsystem and capable of providing a slew rate of 120 T/m/s. This system is equipped with Excite Technology and an 8-channel head coil. In addition, a Signa GE EXCITE 3T short bore MR scanner is available with 40 mT/m gradient subsystem, slew rate of 200 T/m/s and utilizing TwinSpeed dual gradient technology. The 3T MR system is equipped with state of the art 8-channel head coils and allows acquisition of high performance pulse sequences (such as SPGR, FLAIR, DTI) which are necessary for high-quality DTI data acquisition used in this project. The computing facility at the PET/MR Imaging laboratory will be used for the clinical evaluation and testing of the proposed software tool as well as for all data pre-processing steps including PET/MR coregistration, geometric parcellation and integration of quantitative EEG Data with functional PET and DTI data. We will take advantage of both the hardware and software infrastructure already in place. Computational hardware used for this project will include high-end UNIX and PC workstations equipped with specialized software for data pre-processing. This software includes the MEDx (Sensor Systems, Sterling, VA) high performance image processing software, the MPI-Tool (Advanced Tomo Vision, Erfstadt, Germany) coregistration software, the DTIStudio diffusion tensor analysis software (Johns Hopkins University, Baltimore, MD), the 3D-Tool surface rendering software (Max-Planck Inst; Cologne, Germany) as well as in-house developed software routines written in IDL 5.1-6.0 (RSI, Boulder, CO). A secured anonymous ftp server is maintained which allows easy exchange of image data with the Computer Science Department. WSU Department of Computer Science. The Graphics and Visualization Laboratory in the Department of Computer Science will be the main site for development of the proposed software tool due to it's computational resources and theoretical know-how. The 700 square feet facility is specialized in advanced image processing and our development work will take full advantage of the available Pentium-4 PC Workstations and high-end Windows XP PCs as well of the Sun Enterprise 450 workstation which serves as a database server with Oracle 9i installed. If necessary, we will have access to additional infrastructure such as a state-of-the-art SUN Fire 6800 with over 100 processors, several other high-end SUN clusters/servers and access to the University's IBM RS/6000 SP massively parallel computer. Lakeshore Engineering Services, Inc. (LES) will be responsible for the commercial development of our product by providing feedback to the PD and the Co-Is with regard to the requirements of a commercially viable software package. They will assess the market, monitor the position of our product against other medical products, explore marketing strategies and propose changes to the product which will increase its marketability. LES is a Michigan based minority-owned, 8(a), HUBZone, Disadvantaged Business Enterprise (DBE) and a full service multidisciplinary corporation that has been providing complete engineering, environmental, construction, technology & laboratory services to both public and private sectors. LES has offices in Detroit MI, Highland Park MI as well as in other parts of the United States such as St. Louis MO, Omaha NE, Crane IN, Naperville IL, Dover NJ, and Burlington VT. LES has worked successfully with several Federal, State, local and private organizations including the U.S. Army Corps, U.S. Navy, USEPA, GSA, Department of Veteran Affairs, Detroit Housing Commission (DHC), Detroit Public Department (DPW), Public Lighting Department (PLD), Detroit Water Sewerage Department (DWSD), and the City of Highland Park. LES has a team of IT software professionals and project/program managers who are experienced in providing a wide range of support for the commercial development of our product. This support includes commercial network design, business reengineering, E-Business solutions, statistical programming, database design/development, and data warehousing.