Drexel Geometric Biomedical Computing Group

Research           Publications           Personnel

      

Computer Science Department
College of Engineering
Drexel University
3141 Chestnut Street
Philadelphia, PA 19104

      

The Drexel Geometric Biomedical Computing Group conducts research at the intersection of biology, medicine and computer science. The group develops algorithms and software that solve geometry-related computing problems for a variety of biomedical applications.

      

Group Members

      

Collaborators

      

Current Research Projects

      

Interactive, Freeform Editing of Large-Scale, Multiresolution Level Set Models

As imaging and simulation technology continues to be rapidly deployed and utilized in medicine, engineering and science, an increasing number of volume datasets are generated, producing an overwhelming flood of raw 3D data that must be processed, viewed and analyzed. Given the multiple sources of volumetric datasets, there is a great need for powerful implicit model editing capabilities. The aim of this project is to develop techniques and algorithms for interactive freeform editing of large-scale, multiresolution level set models. Level set models combine a low-level volumetric representation, the mathematics of deformable implicit surfaces, and robust numerical techniques to produce a powerful approach to geometric modeling.

The mathematics, algorithms and techniques needed to implement numerous level set modeling capabilities are being developed. Those capabilities include: Freeform Manipulation - Expressive methods for manipulating / modifying the shape of a level set surface, both locally and globally; Topology Control - Allow the user to control if and when intersecting level set surfaces should merge or stay separated; Large-scale - Provide a scalable level set surface data structure that is able to represent complex models containing fine detail; Multiresolution - Allow the user to manipulate/edit the surface at different geometric scales and levels of detail; Interactivity - The user should be able to manipulate/edit the level set surface at interactive (30+ fps) rates.

Preliminary work in this area investigated the effectiveness of tensor voting tokens as geometric modeling constructs.

Related Publications

Museth et al. 2003, Museth et al. 2005, Beltowska et al. 2008, Eyiyurekli and Breen 2009, Eyiyurekli et al. 2009.

J. Beltowska's Research Day 2008 Poster

M. Eyiyurekli's ACM SIGGRAPH Symposium on Interactive 3D Graphics and Games 2009 Poster

      

Simulation of Chemotaxis-based Cell Aggregation and Sorting

The GBC Group is developing a computational model and simulation system that may be used to investigate the fundamental biological processes controlling chemotaxis-based cell aggregation. Chemotaxis is the phenomenon where cells emit chemicals (chemoattractants) into and detect chemical gradients within their environment. They respond to the chemical stimulus by moving in the direction of the chemical gradient and attaching themselves to the other cells upon contact. This process produces cell aggregates that are the building blocks of biological tissue and structure formation. Understanding the influence of the many components of chemotaxis on overall cell aggregation may lead to technologies for tissue engineering based on controlling or directing these underlying biological processes. The key technical challenge in this project is to define a computational model that can be computed efficiently and can effectively capture the biological phenomena of interest, ultimately leading to a predictive simulation capability that supports the discovery of new knowledge in cell biology.

The computational model has been extended to simulate the sorting of heterotypic cell populations. The model for studying cell sorting consists of a subset of features from the chemotaxis-based cell aggregation model. The cells in our sorting experiments do not attach to each other, divide or die, but they do age, and emit and respond to chemoattractant gradients. Additionally new features/parameters have been added to each cell type, namely the excretion of a distinct chemoattractant, a chemotactic response rate and a probability of gradient following for each chemoattractant. Our initial in-silico cell sorting experiments only contained two types of cell populations, T1 and T2. Each cell type emits a unique chemoattractant chemical, C1 and C2 respectively. Both types of cells can sense/respond to both types of chemicals and the strengths of these interactions are defined with parameters L1 and L2. A cell's velocity is proportional to the sum of the sensed, adjusted gradients. Another parameter added in the sorting model is P1 and P2, the probability that a cell will follow the gradient of a specific chemoattractant (C1 and C2) during a simulation time step. If a cell does not respond to a gradient it takes a random step. This computational model produces the sorting results included below.

Related Publications

Eyiyurekli 2006, Eyiyurekli et al. 2007a, Eyiyurekli et al. 2007b, Eyiyurekli et al. 2008.

M. Eyiyurekli's Research Day 2007 Poster

M. Eyiyurekli's Research Day 2008 Poster

This research is performed in collaboration with the Drexel Tissue and Cellular Engineering Group,

Cell aggregation simulation results.

 


Cell sorting simulation results.

      

Self-Organizing Primitives for Automated Shape Composition

Motivated by the ability of living cells to form into specific shapes and structures, the GBC Group is developing a new approach to shape modeling based on self-organizing primitives whose behaviors are derived via genetic programming. The key concept of our approach is that local interactions between the primitives direct them to come together into a macroscopic shape. The interactions of the primitives, called Morphogenic Primitives (MP), are based on the chemotaxis-driven aggregation behaviors exhibited by actual living cells. Here, cells emit a chemical into their environment. Each cell responds to the stimulus by moving in the direction of the gradient of the cumulative chemical field detected at its surface. MPs, though, do not attempt to completely mimic the behavior of real cells. The chemical fields are explicitly defined as mathematical functions and are not necessarily physically accurate. The explicit mathematical form of the chemical field functions are derived via genetic programming (GP), an evolutionary computing process that evolves a population of functions. A fitness measure, based on the shape that emerges from the chemical-field-driven aggregation, determines which functions will be passed along to later generations. We have shown that MPs may be used to define field functions that produce a number of simple shapes. The GP-based process has also generated field functions that produce a number of unexpected, but interesting, patterns and shapes.

Related Publications

Bai et al. 2008a, Bai et al. 2008b, Bai et al. 2008c, Bai 2008, Bai and Breen 2009.

L. Bai's Research Day 2008 Poster

L. Bai's Research Day 2009 Poster

    

    

Morphogenic Primitives self-organizing into ellipse, diamond, hourglass and gear-like shapes.

      

Unexpected and interesting shapes and patterns produced during MP evolution.

      

Shape Distributions for Histologic Grade Estimation

The GBC Group has begun to explore computational methods that may be used to automatically estimate the histologic grade of breast cancer tumors. Breast cancers can be histologically categorized (graded) based upon their architectural patterns and cellular types. Inaccurate histologic grading can result in inappropriate treatment for a given patient. Computational analysis of breast cancers offers an operator-independent method for histologic grading that should enhance grading reliability. Our approach for automatically and objectively estimating histologic grade is based on image processing and shape analysis of imaged histologic sections. Our work is based on the hypothesis that cellular structures found in breast cancer tumors can be transformed into distinct high-resolution shape distributions using geometric measures from stochastic geometry. The resulting shape distributions define well-populated regions of the associated high-dimensional space. Mapping an unknown breast cancer sample into this high-D space and determining to which region it belongs will allow for the automatic estimation of its histologic grade.

Related Publications

Zhang 2008a, Zhang 2008b

J. Zhang's Research Day 2008 Poster

J. Zhang's EMBC 2008 Poster

This research is performed in collaboration with the Drexel's Advanced Pathology Imaging Laboratory.

The radial contact distribution may be used to identify the histologic grade of an unknown sample (far right) by comparing its associated shape distribution with those generated from known samples.

      

Bone Scan Analysis

Drexel's Bone Biology Laboratory studies the microscopic structure of human bones in order to understand the relationship between these structures, aging, health and the large-scale mechanical properties of bone. The Facility acquires high-resolution microscopic optical images of bone under various lighting and filtering conditions, as well as microCT scans.

Cortical bone modeling and drift plays an important role in determining adult bone shape, quantity and quality. With increasing evidence that the bone we grow as children likely affects the health of our bone as we age, it is important to better quantify this process through ontogeny. Moreover, automated means of quantifying these processes can be useful for evaluating the consequences of various health conditions and nutritional deficits on bone growth, important both in modern, orthopedic contexts, and in anthropological contexts to gain insights into the functional adaptations of past populations. In order to investigate this issue the GBC Group is developing a methodology for user-guided segmentation of microCT images of cortical bone cross-sections to (1) discriminate regions of periosteal primary bone apposition that reflect the history of cortical drift in a long bone shaft section and (2) perform measurements that can provide information on bone shape through ontogeny that can then be computationally compared with other the shapes of other bones.

Related Publications

J. DiCristo's STAR 2009 Poster

1) A microCT scan of an adolescent bone. 2) Screenshot of program for specifying segmentation parameters. 3) Results of the segmentation. Pores that fall within a specified size range are highlighted in green. 4) Radial lines are used to calculate the thickness of the periosteal primary bone region (between the circular polyline and the bone boundary) as a function of angle.

      

Contour-based Surface Reconstruction

A wide variety of objects, animals and specimens are scanned for scientific purposes every day in imaging centers across the globe, producing a steady stream of volumetric datasets. Objects such as developing mouse and frog embryos, rat and monkey brains, nerve cells of all types, bones and even fossils are examined by MRI, CT, ET scanners, as well as physically sliced and imaged to produce 3D samplings of these real objects. Once the objects/specimens have been imaged the resulting volume datasets can be manually segmented. In this process, an experienced anatomist goes over selected slices (i.e. images) of the dataset, identifies relevant structures, and circles them with a stylus, producing a series of parallel contours that outline the object of interest. From these sets of contours it is usually required to produce a high quality, smooth 3D surface model that reconstructs the original object. The reconstructed surface is useful for visualization and further processing.

To date, the GBC Group has investigated three solutions to the contour-based surface reconstruction problem. In the first project, which was led by Dr. Ken Museth of Linköping (Sweden) University, we utilize velocity-adjusted 2D level set contour morphing. With this approach, morphing one contour into the next sweeps out a 3D surface. This is accomplished by equating time in the 2D contour morphing process with the third spatial dimension. A critical aspect of this approach utilizes distance estimates corresponding to the arc lengths of trajectories that connect the adjacent contours in the image plane. These distances, together with a time-of-arrival, are used to estimate the speeds (in contour normal directions) needed to produce a smooth morph when transitioning between sets of contours.

The second project investigated the effectiveness of Multi-level Partition of Unity (MPU) implicit models to reconstruct surfaces from noisy input contours. Almost all contour-based surface reconstruction techniques exactly interpolate the input contours. MPU implicit surfaces are a type of point set surface that approximates input data within user-specified error bounds. Thus they offer an approach that inherently copes with noisy data in a controllable fashion. The MPU-based reconstruction technique interprets the contour data (pixels in individual images) as points in 3-space. Since MPU implicits also require normal information, it was necessary to develop an algorithm to estimate surface normals from the stacked contours.

A third technique is currently being investigated that utilizes spline-based 2D distance field interpolation to produce a volumetric representation of the reconstructed surface. A number of filtering techniques are being investigated in order to remove the medial axis discontinuities that are found in distance fields. This third reconstruction approach has proven to be most effective when processing high-complexity, multi-component contours.

Related Publications

Nilsson et al. 2005, Braude 2005, Marker et al. 2006, Braude et al. 2007, Petushi et al. 2008

J. Marker's Research Day 2006 Poster

This research is performed in collaboration with the Graphics Group of Linköping (Sweden) University, the DUCoM Laboratory for Bioimaging and Anatomical Informatics and the DUCoM Advanced Pathology Imaging Laboratory.

      

Contour-based surface reconstruction using 2D level set metamorphosis. Pelvis dataset with 35 input contours.

      

             

Contour-based surface reconstruction using MPU implicit surfaces. Mouse embryo skin dataset with 186 input contours (46,204 points). Mouse embryo heart dataset with 34 input contours (4,528 points). Mouse embryo stomach dataset with 34 input contours (4,088 points).

      

Contour-based surface reconstruction using monotonicity-constrained splines to perform 2D distance field interpolation (48 contours).

      

      

Contour-based surface reconstruction using monotonicity-constrained splines to perform 2D distance field interpolation. Segmented and classified breast cancer histology image. Color-coded 3D reconstruction of the breast cancer tumor from 9 slices.

Breast cancer tumor reconstructions (distance interpolation method)

      

Publications

Dr. Breen's Complete Publication List

Former Students

The GBC Group is financially supported by the National Science Foundation, the Commonwealth of Pennsylvania and Drexel University.

Last modified on August 18, 2009.