The Biomedical Systems Research Group within the Department of Electrical and Computer Engineering is highly interdisciplinary, and maintains active collaborations with other departments and research centres on campus. During the past few years, the ECE Department has hired a number of new faculty members, adding to the core of researchers. The group is comprised by faculty members who are active researchers at the Robarts Research Institute, the Canadian Surgical Technologies and Advanced Robotics research centre, the Lawson Health Research Institute, the Department of Medical Biophysics, and the Faculty of Health Sciences.
Medical imaging techniques such as ultrasound (US), magnetic resonance (MR) and computed tomography (CT) provide detailed 3D images of internal organs. Quantitative information such as organ size and shape can be extracted from these images in order to support activities such as disease diagnosis and monitoring and surgical planning. However, in order to do this, it is first necessary to segment, or extract, mathematical descriptions of organ shape from these images. The descriptions must be accurate and the segmentation process must be repeatable in order to be clinically useful. In addition, in order to compare or combine information from multiple medical imaging modalities, the anatomical scans must be appropriately aligned. The need for image registration often arises from tissue translation, rotation or deformation due to mechanical factors such as body position and tissue compression that, in practice, cannot be precisely controlled while the patient undergoes the imaging procedure. This is a classical problem in the literature, and our researchers are exploring a number of approaches including efficient search methods, statistical combination and classification, maximum entropy methods, and efficient humancomputer interfaces for guiding these automated approaches.
The Cardiothoracic surgical team at the London Health Sciences Centre (LHSC) recently achieved the distinction of being the first in the world to perform a coronary bypass operation on the beating heart using minimallyinvasive intercosal ports and telerobotic equipment. The surgery was performed using a system manufactured by Computer Motion Incorporated, and an Intuitive Surgical Corp. telesurgical robot is also available at the LHSC CSTAR facility. Their “Hermes” console provides a visual display, relayed from an endoscopic camera, which allows the surgeon to perform minimallyinvasive surgical interventions. In this landmark procedure, this system was not equipped with forcefeedback, and a great deal of skill was required for suturing vessels because of the constrained motions imposed by the robot kinematics and the tight workspace of the cardiothoracic cavity. The viewing conditions impose challenges of their own, due to the endoscopic visual field and occlusions from obstructing tissue. Our longterm research objectives are to develop telerobotic system components, and to extend existing telesurgical systems, so that endpoint placement accuracy and elapsed surgical procedure times can be measurably improved. It is anticipated that this research will lead to innovations that can be applied in the domain of minimallyinvasive cardiac bypass surgery, and within the general domain of surgical skills training.
Our specific objectives for research into the development of 3D interactive visualization systems is to adapt existing hardware and software components with the goal of optimising the presentation of visual and haptic information for telesurgical procedures. As an exercise in system prototype design, we are contrasting the bottomup development of system prototypes by domain experts in Biomedical Engineering versus the topdown development of systems by software engineers. Our initial goal is to adapt an existing realtime 3D tracking system to animate a translucent visual model of the beating heart in realtime. This adaptation will require a redesign of the existing software platform to be compatible with realtime programming methodologies, and will lead to recommendations for considering the realtime programming constraints and their impact within an ObjectOriented software design process.
In addition to the measurement problem outlined above, a number of modeling projects are being conducted within the department. In telesurgical planning and control, static preoperative 3D images can only provide a starting point for the planning process. These data must be made more dynamic using realtime measurements. This is a difficult problem as very little is known about the dynamics of internal organs and tissues as they deform during live surgical procedures. Nevertheless, a number of preliminary studies have been initiated; some in conjunction with the Robarts Research Institute and with CSTAR. The use of ultrasound measurements to animate models of internal organs is also under investigation.
In many patients, the quality of diagnostic ultrasound images are limited by focusing errors labelled aberrations. Sound speed variations and scattering from heterogeneous structures located in the outer tissue layers are the principal sources of aberration. Aberration can significantly reduce spatial resolution and contrast when imaging through the breast or abdominal wall. This problem restricts the role of ultrasound in breast cancer screening and complicates the detection of abnormalities in the liver, kidneys, and other abdominal organs. The shortterm goals of this project are to substantially increase the realism of existing algorithms for computations of ultrasonic propagation in tissue and to use the resulting software to develop and evaluate methods for aberration correction. The longterm goal of this project is to create a flexible ultrasonic imaging simulator that can be applied to other important imaging research topics. These simulations will promote the development and testing of new imaging and interventional technologies and will be incorporated into the training of clinical personnel who perform the procedures. Specific active projects include (1) volumetric measurement of tumour progression in mouse models of cancer, (2) identification of the primary cause of early heart failure in a mouse model of cardiac development, and (3) threedimensional analysis of alterations in the layered microstructure of aortic valve tissue that are induced by the processes used to prepare prosthetic valves.
In addition, it is well known that variations in tissue elasticity are associated with the presence of cancer. Manual palpation of the breast and the prostate is a simple qualitative technique used by clinicians for cancer assessment. Elastography has been developed over the past decade as a quantitative elasticity imaging technique to detect breast, prostate and other types of cancer. Over the past few years, novel robust Magnetic Resonance Elastography (MRE) reconstruction techniques have been developed that can be reliably used for breast cancer diagnosis. One obstacle associated with using elastography as a clinical diagnostic tool is its dependence on various parameters used in the elastography procedures. This makes interpreting the resulting elastograms very difficult. Recent work on the development of a nonlinear elasticity reconstruction model which not only addresses this issue but also leads to higher quality and more reliable elastograms.
There are several projects currently underway that deal with improving the runtime efficiency and fidelity of simulation and registration algorithms. In addition, through componentbased ObjectOriented software engineering, these applications are being developed in a way which allows graduate student research from the various interdisciplinary groups to combine their efforts in a structured way. Typically, the humancomputer interface design issues present significant software designs challenges, as the presentation of large datasets of 3D imagery is a crucial portion of the software functionality. Researchers are currently involved in developing a flexible framework for image analysis systems and proposes to address a crucial need in computer vision by encompassing expert knowledge within the system. This has the short term potential to provide a much higher level and more intuitive user interface for a complex image analysis system. Such an interface makes it readily available for nonexpert users allowing widespread use in different domains. In the long term, such a framework will help research machine learning algorithms and such advances would prove crucial in the next generation of intelligent systems.
Currently, the ECE Department has excellent research labs and computing facilities. Since 1999, equipment and software purchases amounting to over 1.71 million have been made for the laboratory and computing facilities in the three programs offered by the Department. In December/2003 the Department Faculty members moved into a much larger space in the new Thompson Engineering Building All established members of the BME group have NSERC research grants. Several members have been successful in securing funding from other sources to establish stateoftheart research facilities.
Our faculty members are members of two overlapping research teams whose collaboration formed the basis for two major grant applications which were subsequently approved. Both of these significant endeavors have lead to the formation of an unparalleled teaching and research infrastructure
Through the first endeavour, the faculty members was part of a team which submitted a proposal for funding to the Whitaker Foundation to support a researchintensive graduate programme in Biomedical Engineering. The proposal was approved for 1,400,000 in order to create new faculty positions in the Faculty of Engineering Science and the Faculty of Medicine and Health Sciences. This grant forms the foundation for a programme to attract new graduate students, and it creates a critical mass of faculty members who will attract excellent students worldwide.
The second endeavour lead to the formation of a research centre, The Canadian Surgical Technologies and Advanced Robotics centre (CSTAR: www.cstar.ca/biographies.asp ), headquartered in a twostory addition which has been construction at the LHSC University Hospital campus. These groups include members of the U.W.O. Faculty of Engineering Science, The John P. Robarts Research Institute, and clinicians at the London Health Sciences Centre (LHSC, University Hospital Campus).
The BME group is one of the most interdisciplinary of research groups. In addition to our regular faculty appointments, the Department of Electrical and Computer Engineering has several Adjunct Faculty members with International Recognition in the BME field, including: Dr. Aaron Fenster, Dr. Terry Peters, Dr. S. Stergiopoulous. In collaboration with the Departments of Medical Biophysics and Anatomical Pathology at Sunnybrook and Women’s College Health Sciences Centre (SWCHSC), breast MR Elastography (MRE) techniques are being developed with the aim of presenting a clinically viable imaging technique for breast cancer assessment. This work is funded by CIHR with Dr. Donald Plewes. Moreover, in collaboration with Dr. Plewes and Dr. Judit Zubovits from the Department of Anatomical Pathology of SWCHSC, measurement of linear and nonlinear elastic properties of ex vivio normal and pathological breast tissues has been carried out through funding from the US Army. In general, BME group members have several active collaborative research projects with industry and with other research centres, e.g.: