Performance validation of nonlinear registration algorithms is a difficult problem due to the lack of a suitable ground truth in most applications. However, the ill-posed nature of the nonlinear registration problem and the large space of possible solutions makes the quantitative evaluation of algorithms extremely
important. We argue that finding a standardized way of performing evaluation and comparing existing and new algorithms currently is more important than inventing novel methods. While there are already existing evaluation frameworks for nonlinear inter-subject brain registration applications, there is still a lack of protocols for intra-subject studies or soft tissue organs. In this work we present
such a framework which is designed in an ”open-source” and ”open-data” manner around the Insight Segmentation & Registration Toolkit. The goal of our work is to provide the research community with the basis framework that should be extended by interested people in a community effort to gain importance for evaluation studies. We demonstrate our proposed framework on a sample evaluation and release its implementation and associated tools to the public domain.

A Framework for Comparison and Evaluation of Nonlinear Intra-Subject Image Registration Algorithms

Open Science includes access to both open source software/methodologies and open data. While there has been progress in open image databases, the results of these efforts are under-reported. As such, imaging scientists are unaware of the available data. In addition, many researchers are interested in providing their data to the greater research community, but may be unaware of the process to release the data. The purpose of this paper is to describe our efforts in developing an open website which includes information on accessible medical image databases as well as some of the logistics for providing an open image database. Most importantly, the authors are requesting participation from the community to contribute to the Medical Image Database Repository (http://midr.org) in order to consolidate the collect knowledge of the community.

Data, data everywhere, nor an image to read - Finding open image databases.

This paper describes the ITK implementation of a transform initializer to produce a rigid 3-D transform able to center a given point on an arbitrary plane. The resultant transform allows the user to "look at" a point from a given direction. The resultant transform can also be used to obtain an oblique section: a single arbitrary 2-D slice through a 3-D image.

LookAt Transform Initializer and Oblique Section Image Filter

White matter tractography enables studies of fiber bundle characteristics. Stochastic tractography facilitates these investigations by providing a measure of confidence regarding the inferred fiber bundles. This article presents a multithreaded ITK stochastic tractography filter that will enable novel studies of fiber tract abnormalities. Additionally, we provide an easy to use command line interface for the filter that is compatible with the 3D Slicer visualization environment.

A White Matter Stochastic Tractography System

As part of the Center of Excellence in Genomic Science at Caltech, we have initiated the Digital Fish
Project. Our goal is to use in toto imaging of developing transgenic zebrafish embryos on a genomic scale
to acquire digital, quantitative, cell-based, molecular data suitable for modelling the biological circuits
that turn an egg into an embryo. In toto imaging uses confocal/2-photon microscopy to capture the
entire volume of organs and eventually whole embryos at cellular resolution every few minutes in living
specimens thoughout their development. The embryos are labelled such that nuclei are one color and cell
membranes another color to allow cells to be segmented and tracked as they move and divide. The use
of a transgenic marker in a third color allows a variety of molecular data to be marked. In toto imaging
generates 4-d image sets (xyzt) which can contain 100,000 to 1,000,000 images per experiment. We are
developing a software package called GoFigure to visualize, segment, and analyze these very large image
sets. GoFigure uses a MySQL database back end for managing storage of images and segmented objects
and uses VTK and ITK for visualization and segmentation. We plan to use in toto imaging to digitize the
complete expression and subcellular localization patterns of thousands of proteins throughout zebrafish
embryogenesis. This genomic data, our zebrafish lines, and GoFigure will be distributed following the
Open Data/Open Source model.

GoFigure and The Digital Fish Project: Open tools and open data for an imaging based approach to system biolgy

This paper describes a new open-source tutorial for image-guided therapy (IGT) and medical robotics that is both accessible and hands-on using the LEGO Mindstorms NXT (a commercially available robotics kit) and 3D Slicer (an open-source application for medical image processing). The tutorial covers all stages of a typical IGT or medical robotics procedure, including the concepts of imaging, preoperative planning, targeting and tracking, navigation and registration, by using the LEGO robot to perform a “needle biopsy” on a phantom (anatomical model) made of traditional LEGO pieces. In addition, this paper describes a C++ library that allows direct control of a LEGO Mindstorms NXT robot from a Linux computer over a USB connection.

An accessible, hands-on tutorial system for image-guided therapy and medical robotics using a robot and open-source software

This document describes work-in-progress for refactoring ITK’s registration methods to exploit the parallel, computation power of multi-processor, shared-memory systems. Refactoring includes making the methods multi-threaded as well as optimizing the algorithms. API backward compatibility is being maintained. Helper classes that solve common registration tasks are also being developed.


The refactoring has reduced computation times by factors of 2 to 200 for metrics, interpolators, and transforms computed on multi-processor systems. Extensive sets of tests are being developed to further test operation and backward compatibility.

Optimizing ITK’s Registration Methods for Multi-processor, Shared-Memory Systems

In this work, 
we extend a previously demonstrated entropy based groupwise registration method
to include a non-rigid deformation model based on B-splines. 
We describe an open source implementation of the groupwise registration algorithm 
using the Insight Toolkit ITK.
We provide the source code, parameters, input and
output data that we used for validation.


We describe an efficient implementation of the algorithm 
by using a stochastic optimization scheme embedded in a multi-resolution setting.
The objective function is optimized using gradient descent algorithm combined
with line search for the step size.
The derivative of the objective function is evaluated efficiently by computing Jacobian of 
B-spline deformation field locally.


We demonstrate the algorithm in application to different imaging modalities
including proton density, FA, T1 and T2 MR images. 
We validate the algorithm on synthetic datasets varying from 2 to 30 images
by recovering randomly applied affine and B-spline transforms.

Non-rigid Groupwise Registration using B-Spline Deformation Model

It is sometimes useful to apply a filter to a region defined by a
mask. This article introduces a simple ``mask adaptor'' filter that
allows this type of operation to be tested easily. It is applicable to filters that do not need any pixel location information, such as histogram operations, which can't necessarily be achieved by simple masking approaches.

Image mask adaptor for pixel order independent filters

In this paper we present a generalization of Frangi's vesselness measure \cite{Frangi1998} for 
the enhacement of M-dimensional shapes in N-dimensional images, together 
with the implementation of the corresponding Insight Toolkit (ITK) filter \cite{ITKSoftwareGuide}. Inspired 
by the implementation of a multiscale vesselness measure recently presented on 
the Insight Journal \cite{Enquobahrie2007}, we also propose a unified framework for the evaluation of 
generic multiscale Hessian-based measures. The manuscript is accompanied by 
source code and examples.