Biomedical Engineering - Carnegie Mellon University

Fancy Numbers

Biomedical Imaging & Signal Processing

The strength of the Department of Biomedical Engineering in this area is reflected by the breadth and depth of research, which leverages a strong computation and biomedical imaging community in the Pittsburgh area. The wide range of research spans from probe development to signal processing.

Development of Biosensors for Cell and Tissue Imaging

Research in this category focuses mainly on developing fluorescence probes for microscopic imaging of live cells and contrast agents for MRI. Through extensive collaborations, these imaging probes have been applied to a broad range of basic and translational studies.

Fluorescent Sensor

Research led by Alan Waggoner, Marcel Bruchez, and Bruce Armitage is aimed at developing new approaches for detecting processes ranging from single molecules to whole animals. The group of Alan Waggoner works on developing fluorescent biosensors to measure enzymatic activities in cell signaling networks (upper figure to the left), while the group of Marcel Bruchez develops fluorescent probes for sensitive single molecular detection. In addition, these groups are collaborating to develop "fluoromodules", which combine a cell-permeable non-fluorescent molecule with an expressed protein to generate fluoresecent biosensors of high photostability, sensitivity, and versatility for direct visualization of dynamic cellular processes (lower figure to the left).
MRI Agent In addition to fluorescent biosensors, new MRI contrast agents have been developed to label protein structures or cells for tracking their localization and fates in animal models. The group of Chien Ho uses sub-micron, functionalized iron particles (figure to the right) to detect the infiltration of immune cells into a transplanted heart during rejection. The group of Michael McHenry uses functionalized magnetic nanoparticles attached to tissue engineering scaffold to image the degradation of scaffold materials following surgical implantation.

Biomedical Image/Signal Processing and Informatics

Drawing on the strengths of Carnegie Mellon University in computation, research in the Department of Biomedical Engineering emphasizes the integration of latest advances in signal processing, computer vision, and machine learning for biomedical applications. Faculty members, including Jelena Kovacevic, Jose Moura, Robert Murphy, Gustavo Rohde, George Stetten, and Jessica Zhang, are working on a broad range of biomedical imaging projects that involve computation.

Bioimaging #10 Jelena Kovacevic leads the research on developing new multi resolution algorithms for segmenting biological and histological images, while the group of Gustavo Rohde works on transport-based morphometry for rigorous classification of biomedical images. Gustavo Rohde has also developed algorithms for image registration, segmentation (figure to the left), and artifact removal for a range of applications in biomedical imaging.
Image Informatics Much of the research in this area emphasizes the “informatics” aspect of image analysis by integrating machine learning and statistical techniques for mining the rich information from complex images. For example, the research of Robert Murphy focuses on developing techniques for analyzing large volumes of data from high-throughput fluorescence microscopy. His group is developing techniques to study complex patterns of protein localization upon cell stimulation (figure to the right).
Cell Tracking Other faculty members working in this area include Lee Weiss, Ge Yang, and Yu-li Wang, who cover a broad range of topics including automatic tracking of cell migration (figure to the left), high-precision measurements of cell motility, and superresolution optical microscopy.
Neural Path In addition to image processing, Biomedical Engineering faculty have been developing and applying signal processing algorithms in the area of computational neuroscience. The group of Byron Yu develops new machine learning techniques to elucidate how large populations of neurons process information, from encoding sensory stimuli to guiding motor actions (figure to the right), while the group of Steven Chase uses statistical analyses to characterize the mechanisms underlying learning and adaptation. Their work takes advantage of powerful neural recording technology that allows simultaneous monitoring of tens to hundreds of neurons.

Applications of Biomedical Imaging Techniques

Hemodynamics Biomedical Engineering faculty, including Kris Dahl, Robert Murphy, Kerem Pekkan, Yu-li Wang, Lee Weiss, and Ge Yang have used image processing approaches in a wide range of studies. A common feature is to combine image data collection and analysis with biological, chemical, physical, or materials techniques to address important biomedical questions. For example, the research of Kerem Pekkan characterizes cardiovascular flow by combining micro particle image velocimetry with fluid mechanics to study embryonic heart development and heart diseases (figure to the left).
Axonal Transport For the research in cell mechanics, the group of Kris Dahl combines quantitative fluorescence imaging with molecular biological techniques to study mechanical properties of the cell nucleus. The group of Yu-li Wang combines fluorescence imaging with micromanipulation and substrate micropatterning to study cell migration and mechanical interactions between the cell and materials surface. The group of Ge Yang applies high-resolution fluorescence imaging to study the mechanics and regulation of intracellular cargo transport in neurons (figure to the right).
Otitis Diagnosis The group of Jelena Kovacevic, in collaboration with the group of Alejandro Hoberman at University of Pittsburgh Medical Center, has developed an image classification algorithm as a diagnostic aid for otitis media. They have defined and distinguished several categories of otitis media to improve diagnostic accuracy (figure to the left).