Biomedical Engineering is inherently an interdisciplinary field. Innovative research involves synergistic collaborations of disciplines in science, engineering, and medicine. The objective of the GWIBE Interdisciplinary Research Fund is to provide impetus for biomedical researchers to seek collaborations across disciplines. We welcome researchers who are active in their own disciplines but have not had the opportunity to participate in biomedical engineering research. Prospective applicants may contact GWIBE for assistance in identifying potential collaborative partners. The ultimate objective of this initiative is to provide seed funds to develop successful externally-funded proposals in biomedical engineering.

Visit here for a list of past winners.



Success of neonatal endotracheal intubation (ETI) is measured by the ability of the operator to place the endotracheal tube in the correct position within 30 seconds. Although it is a critical skill, pediatric trainees often are unsuccessful at its performance. Task trainers and animal models coupled with instructor feedback have been traditionally used to teach neonatal intubation to residents. The small size of the neonatal airway and the large class size and limited time available during training sessions, however, hinder instructors' ability to detect specific causes of procedural failure that can be used as feedback to direct individual learning. Smart tools that enable rapid detection of learner deficits can enhance the learning process by facilitating immediate feedback. The objectives of this pilot study were to: 1. develop a computer based quantitative assessment tool capable of recording and measuring learner performance during simulated neonatal ETI; and 2. compare novice ETI performance to expert performance. A standard neonatal resuscitation mannequin head, laryngoscope and 3.0 endotracheal tube (ETT) were fitted with electromagnetic trackers to capture mannequin head motion and the motion of the laryngoscope and ETT with 6 degrees of freedom. Replica 3D computer models of the head, laryngoscope and endotracheal tube were then developed and registered to align completely with their physical counterparts. All motions were captured and mirrored by the 3D model. Following a warm up period, participants were recorded performing endotracheal intubation three times. Participants recruited included expert neonatology attendings (more-than 60 patient intubations), novice nurse practitioners and novice pediatric residents (less-than 25 intubations). Data was processed and simultaneously sent to a laptop screen for continuous, real-time display of the mannequin, laryngoscope and ETT position and orientation. The software recorded each procedure allowing later review by the instructor.

Participants: Lamia Soghier MD, Wei Li, Rehab Alahmadi, Randall Burd MD, James Hahn


RGB-D scan of the surface of human body for biomedical applications


Physicians use slices and 3D volume visualizations to place a diagnosis, establish a treatment plan and as a guide during surgical procedures. There is an observed difference in 2D and 3D visualization objectives of the various groups of specialists. We describe a generalized temporal focus+context framework that unifies different widely used and novel visualization methods. The framework is used to classify already existing common techniques and to define new techniques that can be used in medical volume visualization. The new techniques explore the time-dependent position of the framework focus region to combine 2D and 3D rendering inside the focus and to provide a new focus-driven context region that gives explicit spatial perception cues between the current and past regions of interest. An arbitrary-shaped focus region and no context rendering are two novel framework-based techniques that support improved planning of procedures that involve drilling or endoscopic exploration. The new techniques are quantitatively compared to already existing techniques by means of a user study.

Participants: Nadezhda Radeva, Lucien Levy, James Hahn
[Published paper]


Tumors that involve the oral cavity, can arise from the tongue, floor of the mouth, lips, and hard palate, among other structures. Advanced tumors can be quite destructive and invade the mandible (lower jaw) or maxilla (upper jaw). Affected segments of the mandible or maxilla are then removed surgically. The standard of care for bony reconstruction of mandibular and maxillary defects involves use of the fibula free flap. The fibula bone is harvested and brought up into the neck with its attached pedicle, the peroneal artery and vein. The peroneal artery and vein are then anastomosed to donor arteries and veins in the neck and the reconstructed bone segment is plated into place. Dental implants can then be placed into the reconstructed mandible at a later time to provide for dental rehabilitation. Depending on the type of defect that is created, wedges of the fibula may need to be osteotomized in the horizontal and/or vertical dimensions to re-create the normal shape of the native patient's mandible. Additionally, skin overlying the lateral aspect of the leg is also harvested along with the fibula if needed. This skin is supplied by 2-3 sub-millimeter perforator vessels coming off the peroneal artery and vein just posterior to the fibula at different points along the length of the peroneal vessels. For these composite defects, osteotomies need to be performed without damaging the perforator vessels to the skin, which are usually not visualized well during the dissection. Ideally the wedge osteotomies are performed in a segment of the fibula bone which is away from the perforator vessels in order to prevent inadvertent injury. Historically after mandibulectomy, for large defects requiring fibular free flaps, the reconstructive surgeon created a mental three-dimensional (3D) picture in his/her mind, and reconstructed the defect without the use of pre-operative templates or guides. Pre-operative planning is crucial to this kind of surgery. Optimizing the surgery parameters is also needed.

Participants: Manal Aassaf, Wei Li, Arjun Joshi, James Hahn
[Published paper]


We propose a novel non-rigid registration method that computes the correspondences of two deformable surfaces using the cover tree. The aim is to find the correct correspondences without landmark selection and to reduce the computational complexity. The source surface S is initially aligned to the target surface T to generate a cover tree from the densely distributed surface points. The cover tree is constructed by taking into account the positions and normal vectors of the points and used for hierarchical clustering and nearest neighbor search. The cover tree based clustering divides the two surfaces into several clusters based on the geometric features, and each cluster on the source surface is transformed to its corresponding cluster on the target. The nearestneighbor search from the cover tree reduces the search space for correspondence computation, and the source surface is deformed to the target by optimizing the point pairs. The correct correspondence of a given source point is determined by choosing one target point with the best correspondence measure from the k nearest neighbors. The proposed energy function with Jacobian penalty allows deforming the surface accurately and with less deformation folding.

Participants: Manal Aassaf, Yeny Yim, James Hahn
[Published paper]


Volumetric datasets are used in many surgical applications. Displaying the 3D information contained in these datasets on 2D screens effectively is challenging, because of loss of information due to projection from 3D to 2D space. Problems such as opaque structures occluding each other and difficulty of perceiving depth of a given pixel complicates these kinds of volume visualizations. Application of effects like transparency and transfer functions to show certain anatomical structures makes the mental mapping between the visualization and the patient difficult. We have used the focus+context visualization paradigm to help with these problems. The users can explore the datasets by using different transfer functions in different parts of therendering. This way, internal structures can be displayed by exploring multiple co-registered datasets in a single coherent view. A novel rendering method using depth images minimizes the performance impact of this visualization approach.

An extension to this visualization idea is also developed by using volumetric brushes to perform volume editing tasks. This way, the users can perform accumulated selection of arbitrary shaped regions, and combine multiple datasets in a single visualization.

[More info and video] [ICG page]

Participants: Can Kirmizibayrak, James Hahn

  Traditional tactile interaction methods such as trackers or the mouse can cause problems when used in the operating room. Trackers can be expensive, difficult to setup and calibrate and prone to errors due to line-of-sight requirements or electromagnetic interference. The mouse is ubiquitous, but it is a 2D interface and might be difficult in 3D interaction tasks necessary for volume visualization. The common problem for all these interfaces is that they require sterilization, a process that is time consuming, costly, and possibly can cause complications.

The use of gestures can replace these interaction methods and eliminate these problems. Furthermore, intuitive methods like these can decrease the training time necessary for performing interaction tasks important for medical volume visualization. We are developing such methods for common tasks such as data exploration, editing and rotation. The user studies conducted showed that novice users can perform tasks such as matching rotations and finding internal structures in a volumetric dataset very effectively and quickly using a gesture-based interface compared to the mouse.

[More info and video] [ICG Page]

Participants: Can Kirmizibayrak, Nadezhda Radeva, James Hahn



In cancer medicine, the most common predictive factors and prognostic approach is the categorical system of TNM (Tumor, Lymph Node Status, and Metastasis) that defines the anatomic extent of disease at diagnosis. However, practice by medical professional specialties over many years has shown that TNM is useful but needs to be expanded to accommodate additional clinical, diagnostic, and therapeutic responsive predictive factors for higher accuracy in prediction. For example, by 1995, 76 predictive factors had been reported for breast cancer. However, not one of these factors has been incorporated into the TNM. The integration of new predictive factors into comprehensive systems with greater predictive power than the TNM will clearly benefit patients and physicians and is needed.

Currently we are working on novel clustering approaches that predict the outcome of cancer patients. These approaches aim at producing a prognostic system that provides descriptive statistics to ascertain which patients constitute a set of subgroups such that patients within each group are more similar in survival than patients from different groups. While our research focuses on cancer, we should emphasize that this research has general application. These are completely new approaches since it is the first time censored data has been taken into account within the framework of cluster analysis.

From a broader view, we primarily focus on development and application of machine learning techniques (e.g., support vector machines, the method of stochastic discrimination, etc.) including their feasibility and application to laboratory research and patient care. More specifically, we have five major research thrusts: 1) analysis of array data with application to cancer etiology, treatment, and personalized medicine, 2) evaluation and identification of risk factors for cancer and methods of evaluation, 3) clinical outcomes and survival prediction, 4) demographic and epidemiologic analysis, and 5) development and testing of novel computer based algorithms.

For more information, please contact Prof. Susan Cheng at

  Since natural tissue is mainly nanometer in dimension and various cells directly interact with nanostructured extra-cellular matrices, the biomimetic features and excellent physiochemical properties of nanomaterials play a key role in guiding various tissue repair and regeneration. Our Nanomedicine and Tissue Engineering lab directed by Prof. Lijie Grace Zhang applies a range of interdisciplinary technologies and approaches in nanotechnology, tissue engineering, biomaterials, stem cells, drug delivery and biomechanics to create biologically inspired tissue scaffolds and investigate their efficacy for healing various tissues in vitro and in vivo. Specifically, there are three research directions in our lab:
- Biomimetic Nanomaterials for Bone & Cartilage Tissue Engineering and Orthopedic Applications
- Stem Cell Therapy for Tissue Regeneration
- Sustained Drug Delivery Systems for Biomedical Applications
More information can be found via
  (A) Fractured bone
(B) Endothelial cell growth on titanium with nanocoatings
(C) Biologically inspired DNA based rosette nanotube network
(D) The multipotential mesenchymal stem cell for bone and cartilage repair

Ventricular fibrillation (VF), a lethal cardiac arrhythmia, claims the life of almost one thousand Americans each day. A goal of GW's CERL is to understand the electrophysiologic events that cause VF and to investigate ways of preventing it. Computational models of cardiac tissue and innovative instrumentation developed at the CERL are currently used to study how the occlusion of a coronary artery may initiate VF, to improve electrical therapies for VF, and to understand how stem cell engraftment may improve the function of a diseased heart. The CERL is located at the GW Medical Center and is a multidisciplinary effort between the GW School of Engineering and Applied Science and the GW School of Medicine and Health Sciences.


Mechanism of a cardiac arrhythmia resulting from reperfusion after local ischemia.

Left: The time course of local ischemia was monitored as an increase in the epicardial fluorescence of NADH (lower panel) from a rat heart. No arrhythmia was detected before reperfusion (upper panel). Immediately upon reperfusion a tachyarrhythmia was generated and NADH fluorescence began to fall to pre-ischemia levels.

Right: Fluorescence images of transmembrane potential during reperfusion. Initially a fast ectopic (point) source motivated the arrhythmia and it later evolved into a transient reentrant source (rotor).


  It's commonly known there are three states of matter: solid, liquid and gas, but there are actually four. And the last one would be plasma. It's wonderful substance, ionized gas that can be used as a source of the energy of various origins: thermal, electrical or light.
There are variety of different kinds of plasmas ranging from stars in universe to fluorescent lamps or plasma TVs. Recently it was breakthrough research in the area of cold atmospheric plasmas opened the gates to a world of technology straight out of science fiction.
A wide range of cold plasmas applications have been investigated including sterilization (food, medical equipment, contaminated civilian and military gear), the preparation of polymer materials for medical procedures, wound healing, tissue or cellular removal, surgery, treating skin disease and dental drills. Thus Dr. Michael Keidar and his group in collaboration with GW Medical Center are working in biomedical engineering area of studying various aspects of cold plasmas applications, particularly interaction of cold plasmas with living tissue at cellular and subcellular levels (cell adhesion, cell migration ect) and nanomedicine. Also plasma jet analysis (physical and chemical characterization, spectroscopy) and simulations are part of the ongoing work.

  All swimming techniques in competitive swimming are primarily oriented towards two goals: maximizing thrust and minimizing drag. Two factors make these goals difficult to achieve:

1. The water flow past a swimmer, which ultimately determines the total force on the swimmer, is highly complex and turbulent, and little regarding the dynamics of this flow is currently understood.

2. The typical swimming stroke is generally quite complex and composed of a sequence of coordinated body motions. Variations are possible at every stage in the stroke making it difficult to determine which particular composition is optimal in terms of thrust production and drag reduction.

In the past, research in swimming techniques has been limited to relatively simple experiments, theoretical modeling and videographic analysis. However, with the recent advances in computer hardware and software, techniques such as computational fluid dynamics (CFD) and computer animation and visualization can now be brought to bear on this problem. We are working with USA Swimming to systematically addresses both of the above factors.
The near term objective of this research is to produce tangible results that can impact the performance of Team-USA in the 2008 Olympics. In the long term, this research is expected to generate research and development that will greatly improve our fundamental understanding of the fluid dynamics and biomechanics of swimming and produce tools that can be used for coaching, and assessing swimmers at all levels of performance.

Project Poster | High Resolution (3072x2034, TIFF file) | Low Resolution (1024x768, JPEG File)

Demo Video | (.AVI 51.8M) | (.WMV 21.3M)

Also visit:
Computer Animation and Visualization of Olympic Swimmers
Full Body Analysis of Swimming Techniques

  Vocal cord paralysis and paresis are debilitating conditions leading to difficulty with voice production. Medialization laryngoplasty is a surgical procedure designed to restore the voice in patients by implanting a uniquely configured structural support lateral to the paretic vocal fold through a window cut in the thyroid cartilage of the larynx. Currently, the surgeon relies on experience and intuition to place the implant in the desired location, therefore it is subject to a significant level of uncertainty. Window placement errors of up to 5mm in the vertical dimension are common in patients admitted for revision surgery. The failure rate of this procedure is as high as 24% even for experienced surgeons. An intraoperative image-guided system will help the surgeon to accurately place the implant by superimposing the CT data from the patient with the actual larynx of the patient during surgery.
One of the fundamental challenges in our system is to accurately register the preoperative 3D CT data to the intraoperative 3D surfaces of the patient. Our proposed image guided system will use the anatomical and geometric landmarks and points to register intraoperative 3D surface of thyroid cartilage to the preoperative 3D radiological data. The proposed approach has three phases. First, the laryngeal cartilage surface is segmented out from the preoperative 3D CT data. Second, the surface of the exposed laryngeal cartilage during the surgery is reconstructed intraop-eratively using stereo vision and structured light based surface scanning. Third, the two geometries are registered using ICP based shape matching. The proposed ap-proach has several advantages over alternative approaches: the combination of stereo vision and structured light surface scanning is capable of tracking the fiducial markers, reconstructing the surface of laryngeal cartilage and matching the preoperative and postoperative surfaces for registration purposes. The computer vision based approach can be applied to delicate areas like laryngeal cartilage with no danger of causing physical damage.

  We have started a new line of research associated with the use of medical informatics in a number of applications. In one of these applications, we are interested in creating a prototype system capable of detecting in real time, biologic, chemical, nuclear, and radiological terrorism attacks as well as routinely supporting the local public health community with information regarding natural outbreaks of contagious and traumatic diseases. This can be accomplished through syndromic surveillance of the Emergency Medical Dispatch (EMD) interrogatory data, collected by the computer aided dispatch (CAD) system at 911 Centers already in place. A set of tools will allow further analysis to determine possible causes (e.g. type of biological agents), the mechanism of transmission (for example, based on current atmospheric conditions and simulations of the dispersion of the plume), and possible actions (e.g. evacuation of a population center in danger). The system is intended to be used by local, regional, and national public health officials giving them advance information about possible man made or naturally occurring events, and helping them make decisions on possible actions to take.
  The objective of this interdisciplinary research is to give the surgeon the tools to improve the outcome of the procedure and to reduce the revision rate. The research is composed of two interrelated components. The first component is a computer model for simulating vocal fold function during phonation. This will lead to a pre-operative planning system that will guide the surgeon in determining the optimal location and size/shape of the implant. This will be accomplished by developing a patient-specific model of the laryngeal cartilage and vocal folds (with implant included) and simulating the fluid dynamics and vocal fold vibration associated with voice production. The second component comprises an intra-operative image-guided system that will be developed to allow the surgeon to accurately place the implant at the desired location. Image guidance will be performed by registering a pre-operative 3D CT image with the 2D images of the patient during the operation. A novel graphical interface system will allow the surgeon to easily use the system without any major changes to current surgical practices.

The near-term goal of the proposal is to improve the clinical outcomes of the particular procedure. However, the long-term significance of the proposal is to solve fundamental scientific problems associated with the biomechanical modeling and simulation of voice production as well as the relatively unexplored problems associated with the use of intraoperative 2D imagery in image-guided surgical procedures.

  Medical simulations are used for a number of purposes. One of the most promising is the development of surgical simulators. The accepted paradigm for teaching in medicine has been "see one, do one, teach one". Although the methodology has served medicine well, there is a growing interest in the use of computer-based surgical simulators to teach complex surgical procedures. This has been prompted by the prevalence of "minimally-invasive" procedures. Minimally-invasive procedures involve the use of imaging techniques (MRI, CT, ultrasound, laparoscopes) to guide instruments through a small opening in the patient to perform certain surgical procedures. The benefit is the reduced amount of trauma to the patient. However, the procedures have become extremely complex making effective training critical. Personnel from SEAS and SMHS have been involved in a number of research projects that involve computer scientists, electrical engineers, mechanical engineers, and physicians to develop virtual reality simulators that allow physicians to see as well as feel the simulated procedure. This type of training, although of limited use currently, has a great deal of potential in revolutionizing medicine in much the same way that flight simulators have revolutionized pilot training. Similar technology can be used to guide physician to place an instrument (e.g. biopsy needle) at a precise location on the patient using imaging techniques along with computer guidance. We intend to leverage our current expertise in these areas to continue developing sophisticated simulators with the intention of expanding their use in training physicians as well as in the operating room.
  Atherosclerosis is a major cause of death in the United States. Stroke, one consequence of this occlusive disease in a carotid artery, remains one of the leading contributors to major morbidity and mortality within the United States, in which approximately 700,000 new or recurrent cases are encountered annually (AHA 1999, Stroke & CVD 1992). Stroke also has become the leading cause of serious, long-term disability in the United States, accounting for more than half of all patients hospitalized for acute neurologic disease (AHA 1999). Clinicians rely heavily on the imaging modalities, such as Duplex US, in making judgements regarding the severity and need for intervention. Accurate correlation between post-stenotic flow characteristics and stenotic severity will aid immeasurably in making such critical decision-making more reliable. Profs. Guo, Mittal and Hahn will collaborate on the development of acoustic sensors that can provide accurate non-invasive assessment of peripheral vascular disease. Once a correlation between the stenotic severity and bruit frequency has been extracted from the simulations, we will develop a device that will employ this correlation to predict the severity of the arterial disease. The team expects to work closely with researchers from the GW School of Medicine in both the development and the eventual testing of this device.

In addition to this, the team also plans to work on analyzing the hemodynamic origins of "Korotkoff" sounds which form the basis for the indirect, auscultatory method of measuring blood pressure. Despite the availability of highly accurate invasive procedures, this method is still the cheapest and most convenient way of accurately measuring blood pressure. However, in addition to the 10% estimated inaccuracy of the auscultatory method, false measurements are common. The team proposes to use computer simulations to understand the hemodynamic mechanisms that cause these sounds and correlate more precisely the various stages in these sounds to the measured blood pressure. It is expected that such a study could eventually lead to devices ("smart" stethoscopes) that can improve the accuracy of the auscultatory method.

  800 22nd ST, NW Suite 5830
Washington, DC 20052