In this project we will focus on four important aspects (requirements) of an end-to-end software environment for performing multiple parallel non-rigid registrations over the Grid. Our application is Image Guided Neurosurgery (IGNS). IGNS is an important tool for neurosurgical resection which is a common therapeutic intervention in the treatment of malignant brain tumors.
The four requirements we will study in this project are: (1) the high-performance of the parallel and distributed non-rigid registration code by dynamically balancing the processors work-load, (2) the ease-of-use from medical doctors (radiologist) of a web-service by developing a Graphical User Interface (GUI) for clients of our web-service (3) the portability of the code to run from a single high-end workstation to more than one clusters of workstations within the local and wide area networks available to research hospitals and medical centers, (4) the reliability by guaranteeing that part of the system will complete the computations for the image registration and by making sure we meet the time constrains by acquiring and re-configuring the system at runtime to compensate for lost (faulty) computing resources.
The end-product will be a prototype of an IGNS web-service. However, this project will have broader impact to image guided therapy because the same framework (but different modalities and finite element approximation models) can be used in prostate biopsy and treatment, and liver cryotherapy.
The basic research of this project contributes into a broader goal for Research and Development (R&D) efforts for developing enabling technologies for medical applications and devices required in personalized medicine for brain cancer, Parkinson’s and Alzheimer’s disease. We achieved our objective to develop an accurate, faster, and cheaper core technology for fusion (or registration) of medical images required in Image Guided Neurosurgery (IGNS) for brain cancer using commodity computing resources. In addition, we started the evaluation of this technology for Deep Brain Stimulation (DBS) required for the treatment of Parkinson’s disease. We delivered an open source software as part of the Insight Segmentation and Registration Toolkit (ITK). ITK is an open-source, cross-platform system that provides developers with an extensive suite of software tools for image analysis. ITK is funded and supported by NIH’s National Library of Medicine and it is developed through extreme programming methodologies, it employs leading-edge algorithms for registering and segmenting multidimensional data –it is used worldwide. Our paper/software (within less than two years is viewed more about 9,000 times and downloaded about 1,400 times). The non-rigid registration method and software we delivered from the basic research of this grant is : (A) more accurate than the current technology used in commercial products (in average more than 3.5 times) and helped us to improve the accuracy of the state-of-the-art method which is part of the ITK by at least a factor of two, (B) fast enough to meet the time constraints imposed by the neurosurgical procedure and (C) it runs on widely available desktops which are portable and easy to integrate in traditional Operating Rooms. Hospital adoption rate, for devices that rely on enabling core technologies like the non-rigid registration method we developed in this project, will increase as the cost of intra-operative technologies decreases and the employment of such technologies will reduce medical and hospitalization expenses. In addition DBS and IGNS procedures in general are expected to increase, since incidents of Parkinson’s, Alzheimer’s, and brain cancer are expected to increase with the aging of the US population. Additional broader impact is expected to come from applications in the prostate, liver, and kidney -all under evaluation at top research hospitals.
