An Insight Toolkit (ITK) processing framework for blob segmentation with applications to microscopy 
images is presented in this paper. Our algorithm is a reﬁnement of the work of Mosaliganti et al. [3] for 
splitting cell clusters. The basic idea is to incorporate as many cues as possible into developing a suitable 
level-set speed function so that an evolving contour exactly segments a cell/nuclei blob. We use image 
gradients, distance maps, multiple channel information and a shape model to drive the evolution. The 
framework consists of a linear pipeline of 6 new ITK ﬁlters which are applied in succession to generate 
the segmentation. The ﬁlters extract the cell foreground, construct the speed image, ﬁnd seed points for 
evolution and collect cell statistics from the segmentation. We include 2D/3D example code, parameter 
settings and show the results generated on confocal images of the zebraﬁsh embryo.

Microscopy Image Analysis: Blob Segmentation using Geodesic Active Contours

An Insight Toolkit (ITK) processing framework for segmentation using active contours without edges 
is presented in this paper. Our algorithm is based on the work of Chan and Vese [1] that uses level- 
sets to accomplish region segmentation in images with poor or no gradient information. The basic idea 
is to partion the image into two piecewise constant intensity regions. This work is in contrast to the 
level-set methods currently available in ITK which necessarily require gradient information. Similar to 
those methods, the methods presented in this paper are also made efﬁcient using a sparse implementation 
strategy that solves the contour evolution PDE at the level-set boundary. The framework consists of 6 
new ITK ﬁlters that inherit in succession from itk::SegmentationFilter. We include 2D/3D example 
code, parameter settings and show the results generated on a 2D cardiac image.

Level Set Segmentation: Active Contours without edge

An Insight Toolkit (ITK) processing framework for simultaneous segmentation of multiple objects using 
active contours without edges is presented in this paper. These techniques are also popularly referred to 
as multiphase methods. Earlier, we had an implemented the Chan and Vese [1] algorithm that uses level- 
sets to accomplish region segmentation in images with poor or no gradient information. The current 
work extends that submission to use multiple level sets that evolve concurrently. The basic idea is to 
partion the image into several sets of piecewise constant intensity regions. This work is in contrast to 
the level-set methods currently available in ITK which necessarily require gradient information and also 
necessarily segment a single object-of-interest. Similar to those methods, the methods presented in this 
paper are also made efﬁcient using a sparse implementation strategy that solves the contour evolution 
PDE at the level-set boundary. This work does not introduce any new ﬁlter but extends the earlier 
submitted to ﬁlters to process multiple objects. We include 2D/3D example code, parameter settings and 
show the results generated on a 2D cardiac image.

level set segmentation using coupled active surfaces

 Junctions of tubular structures (vasculature, trachea, neuron, etc) in medical images are critical for the topology of these structures. Identification of them is helpful in many applications. For example, quantification of geometric vascular features, registration of trachea movement due to respiration, tracing of neuron path. However, manual extraction of junctions can be tedious, time-consuming, and subject to operator bias. In this paper, we propose a novel method to detect them automatically and describe how to implement it in ITK framework. The input is a 2D/3D binary image that can be obtained from any segmentation techniques. The output will be positions of junctions and their sizes. There are only two parameters which need to be set by the user. We provide here the implementation as a ITK class: itk::JunctionDetectionFilter.
Please cite the following paper if you are interested in our work.
G. Xiong, C. Chen, J. Chen, Y. Xie, and L. Xing, Tracking the Motion Trajectories of Junction Structures in 4D CT Images of the Lung, Vol. 57, No. 15, pp. 4905-4930, Physics in Medicine and Biology, 2012.

Automatic Junction Detection for Tubular Structures

Mathematical modeling and computer simulation have proved tremendously
successful in engineering. One of the greatest challenges for mechanists is
to extend the success of computational mechanics to fields outside
traditional engineering, in particular to biology, biomedical sciences, and
medicine. The proposed workshop will provide an opportunity for
computational biomechanics specialists to present and exchange opinions on
the opportunities of applying their techniques to computer-integrated
medicine. For example, continuum mechanics models provide a rational basis
for analyzing biomedical images by constraining the solution to biologically
reasonable motions and processes. Biomechanical modeling can also provide
clinically important information about the physical status of the underlying
biology, integrating information across molecular, tissue, organ, and
organism scales. The main goal of this workshop is to showcase the clinical
and scientific utility of computational biomechanics in computer-integrated
medicine.

The authors of the best papers presented at the Computational Biomechanics
for Medicine workshop will be requested to submit an extended version of
their papers to the special section of the Medical Image Analysis journal.

Computational Biomechanics for Medicine III: a MICCAI 2008 satellite

We propose a new nonlinear image registration model which is based on nonlinear elastic regularization and unbiased registration. The nonlinear elastic and the unbiased regularization terms are simplified using the change of variables by introducing an unknown that approximates the Jacobian matrix of the displacement field. This reduces the minimization to involve linear differential equations. In contrast to recently proposed unbiased fluid registration method, the new model is written in a unified variational form and is minimized using gradient descent. As a result, the new unbiased nonlinear elasticity model is computationally more efficient and easier to implement than the unbiased fluid registration. The unbiased large-deformation nonlinear elasticity method was tested using volumetric serial magnetic resonance images and shown to have some advantages for medical imaging applications.

Nonlinear Elastic Registration with Unbiased Regularization in Three Dimensions

Nonlinear finite element methods are described in which cyclic organ motion is implied from 4D scan data. The equations of motion corresponding to an explicit integration of the total Lagrangian formulation are reversed, such that the sequence of node forces which produces known changes in displacement is recovered. The forces are resolved from the global coordinate system into systems local to each element, and at every simulation time step are expressed as weighted sums of edge vectors. In the presence of large deformations and rotations, this facilitates the combination of external forces, such as tool-tissue interactions, and also positional constraints. Applications in the areas of surgery simulation and minimally invasive robotic interventions are discussed, and the methods are illustrated using CT images of a pneumatically-operated beating heart phantom.

Inverse Nonlinear Finite Element Methods for Surgery Simulation and Image Guidance

In this paper, we propose a Mix-resolution Bone-related Statistical Deformable Model (mBr-SDM) to improve the predicting accuracy of orthognathic surgery, particularly for the main deformation region. Mix-resolution Br-SDM consists of two separate Br-SDM of different resolutions: a high-resolution Br-SDM which is trained with more samples to capture the detail deforming variations in the main deforming regions of interest, together with a low-resolution Br-SDM which is trained with a smaller number of samples to capture the major variations of the remaining facial points. The experiments have shown that the mix-resolution Br-SDM is able to significantly reduce the predicting error compared with the corresponding Finite Element Model, while giving a low computational cost which is characteristic of the SDM approach.

A Mix-resolution Bone-related Statistical Deformable Model (mBr-SDM) for Soft Tissue Prediction in Orthognathic Surgery Planning

In this paper we propose the usage of a very efficient contact implementation for modeling the brain-skull interaction. This contact algorithm is specially design for our Dynamic Relaxation solution method for solving soft-tissue registration problems. It makes possible the use of complex biomechanical models which include different nonlin-ear materials, large deformations and contacts for image registration. The computational examples prove the accuracy and the computational efficiency of our methods. For a model having more than 50000 degrees of freedom, a complete simulation can be done in less than a minute.

Realistic And Efficient Brain-Skull Interaction Model For Brain Shift Computation

We present a framework for statistical finite element analysis allowing performing statistical statements of biomechanical performance of peri-articular implants across a given population. In this paper, we focus on the design of orthopaedic implants that fit a maximum percentage of the target population, both in terms of geometry and biomechanical stability. CT scans of the bone under consideration are registered non-rigidly to obtain correspondences in position between them. A statistical model of shape is computed by means of principal component analysis. A method to automatically propagate standardize fractures on the statistically-based bone population has been developed as well as tools to optimize implant position to best-fit the bone surface. Afterwards, finite element analysis is performed to analyse the biomechanical performance of the bone/implant construct. The mechanical behaviour of different PCA bone instances is compared for tibia representing the Asian and Caucasian populations.