The goal of radiation treatment is to irradiate cancer cells (i.e., a target region) without destroying adjacent healthy tissue. Thus, it is advantageous to form the beam so that it best approximates the target, thereby reducing the amount of dose absorbed in critical regions outside the target area. While multi-leaf collimators are common in human clinical systems, small animal radiotherapy systems are typically limited to a set of fixed-size collimators. For these systems, dose painting can be used for conformal dose delivery, but is significantly slower than a multi-leaf collimator. As a compromise solution, a variable rectangular collimator has been developed for the Small Animal Radiation Research Platform (SARRP). This enables more efficient dose painting via the decomposition of a 2D target region into a minimum number of rectangles of variable size, which is the topic of this paper. The proposed method consists of several distinct steps and was implemented on the SARRP Treatment Planning System (TPS).

Dose Painting with a Variable Collimator for the Small Animal Radiation Research Platform (SARRP)

Patient motion during radiotherapy or intra-fractional motion, is one of the major sources of uncertainty in dose application. 2D/3D registration is an intensity based method used successfully to track tumor motion with the potential to reduce uncertainty. Despite promising results, validation remains a problem due to the lack of a gold standard. In this paper, we demonstrate the use of clinical datasets acquired with the VERO LINAC to validate intensity based 2D/3D registration. The patient considered in this study had a gold marker implanted to enable live tumor tracking. We verify our results against the tracked gold marker position. Our results show that it is possible to accurately track the tumor using only intensity based 2D/3D registration. The mean error of our registration is of 1.6 ± 0.3mm and as low as 1.1 ± 0.5mm when looking only at the cranial-caudal direction. The results pave the way for a possible clinical application of the method.

Validation of real-time intensity based 2D/3D registration for image guided radiotherapy

Automated segmentation is a frequently applied task in the course of medical imaging. Furthermore, it is a substantial component of image-guided radiotherapy. Atlas based segmentation is one of the most frequently used approach for automated segmentation. Especially for multi-atlas based segmentation, segmentation quality and speed largely depends on the underlying registration and atlas selection strategy. In this work an atlas selection strategy that is based on the correlation of inter-atlas similarities within a set of atlas images is presented. Segmentation quality is analyzed by calculating dice coefficients and 95% Hausdorff distances for the left and right parotid with respect to different numbers of atlases. Results are compared to other state of the art atlas selection strategies. It can be shown that the developed atlas selection technique performs slightly better than NMI-based selection if a low number of atlases is used.

A novel atlas-selection approach for multi-atlas based segmentation using the correlation of inter-atlas similarities

Glioblastoma (GBM) is the most common type of primary
brain tumor, which is characterized by an infiltrative growth pattern.
In current practice, radiotherapy planning is primarily based upon T2
FLAIR MRI despite its known lack of specificity in the detection of tu-
mor infiltration. While hyperintensity on T2 FLAIR is widely considered
to represent infiltrative tumor, it may also be caused by the presence
of vasogenic edema (VE), caused by a leakage of fluid into the brain
parenchyma. Distinguishing VE from infiltrative tumor could have im-
pact on improving radiotherapy planning. In this paper we study a data
set of 17 GBM patients treated with anti-angiogenic therapy for which
a fast decrease of T2 FLAIR hypersignal is observed, which indicates
the resolution of VE. We investigate if multimodal MRI acquisitions in-
cluding diffusion tensor imaging can distinguish between VE and tumor
infiltration prior to therapy. Using a random forest classifier, we show
that, in this study, morphological information based on the contrast en-
hanced T1 image explains up to 75% of the extent of VE. The information
from different imaging modalities did not significantly improve the clas-
sification. We then show that delineating the VE prior to therapy can
have substantial impact on radiotherapy target delineation, leading to
smaller treatment volumes and reducing potentially harmful radiation
dose to normal brain tissue.


Multimodal Analysis of Vasogenic Edema in Glioblastoma Patients for Radiotherapy Planning

Introduction. Computed tomography (CT) is the standard imaging modality for radiation therapy treatment planning (RTTP) because of its ability to provide information on electron density. However, magnetic resonance (MR) imaging provides superior soft tissue contrast, especially in small animal imaging, facilitating the precise selection of the target volume. This makes the technique interesting for irradiation of brain tumors. The aim of this study was to present an MR-only based workflow for RTTP on a small animal radiation research platform (SARRP) by investigating the potential of probabilistic classification of voxels using multiple MR sequences.
Methods. Six female Fisher rats were anesthetized using isoflurane and individually fixed on an in-house made multimodality bed before starting MR and CT acquisitions. MR measurements were performed on a 7-Tesla system using a rat brain volume coil. Four different MR sequences were acquired for each animal, including a T1-weighted (MDEFT) sequence, a T2-weighted (RARE) sequence, an ultra-short echo time sequence with 20 μs echo time (UTE1) and an ultra- short echo time sequence with 2 ms echo time (UTE2). UTE offers the opportunity to acquire images from proton-poor structures with very short transverse relaxation times, such as bone, by using a rapid readout of the fast decaying signal. Following MR, the animals were moved to the SARRP to start a cone-beam CT (CB-CT) by acquiring 720 projections over 360°. Cone-beam CT projection data were reconstructed by filtered back-projection to obtain the standard-CT for RTTP. Then the images were bias field corrected and manually co-registered to the CB-CT. After that, images were segmented in three tissue classes (air, soft tissue and bone) with k-means for the CB-CT and fuzzy c-means segmentation algorithm (FCM) for the MR images with multiple MR images as input. The membership probability can be between 0 and 1, with one indicating 100% probability and zero indicating 0% probability to belong to a specific tissue class. To obtain a pseudo-CT image, voxels were assigned to the tissue class having the highest membership probability. The dice coefficient was used to evaluate the correctness of the segmentation for soft-tissue and bone. The pseudo-CT images with the highest similarity index were used for further radiotherapy treatment planning (RTTP), in addition, to the standard UTE1-UTE2. The target of the RTTP that was selected in the primary cortex (M1) and three different beam arrangements were investigated to compare CB-CT and MR-based dose calculations. The dose plans were a single static beam of 3x3 mm, using a single arc (3x3 mm beam size, 120° arc, couch at 0°), and three non-coplanar arcs (3x3 mm beam size, 120° arc, couch at 0°, 45° and 90°). Dose distributions were calculated using the TPS of the SARRP and cumulative dose volume histograms (DVHs) of the target and normal brain tissue were obtained for the three dose plans.
Results The highest dice coefficient was obtained for the T1-UTE1-T2 combination, which was used for further RTTP. The contribution of bone to the total dice coefficient did not exceed 27%. However, bone accounts for only 2% of the image, therefore a misclassified bone pixel has a bigger effect in the dice coefficient than a misclassified soft tissue pixel. Using only 1 beam, both MR combinations underestimate the dose to be delivered to the target. When more complex beam configurations were used to irradiate the target, very small differences were observed between CB-CT and MR based dose calculations.

Pseudo-CT generation from multiple MR images for small animal irradiation

The contribution to VTK presented in this article is specialized on the extraction of contact surfaces (CS). This extraction can be regarded as the intersection Boolean-operation of only touching meshes. The vtkCoplanarSurfaceExtractor filter produces either polygonal or triangulated CSs by reconstructing the contact faces of co-planar triangles. Specified tolerances account for discrepancies in coplanarity of faces which might occur due to rounding effects.
This article is accompanied with the source code, input data, parameters and output data that were used for validating the VTK-filter.


Extracting Intersections of Coplanar Surfaces (Boolean-operation on touching meshes)

We present a generic interpolator for label images. The basic idea is to interpolate each label with an ordinary image interpolator, and return the label with the highest value. This is the idea used by the itk::LabelImageGaussianInterpolateImageFunction interpolator. Unfortunately, this class is currently limited to Gaussian interpolation. Using generic programming, our proposed interpolator extends this idea to any image interpolator. Combined with linear interpolation, this results in similar or better accuracy and much improved computation speeds on a test image.

A generic interpolator for multi-label images

The presented ParaView plugin allows easy access to the algorithm described in Ref 1. It enables analysis of faceted objects that exhibit distortions in their digital representation, e.g. due to tomographic reconstruction artifacts.
The contributed functionality can also be used outside ParaView in e.g. command-line programs. The code, data, a test and an example program are included.

Facet Analyser: ParaView plugin for automated facet detection and measurement of interplanar angles of tomographic objects

Nonlinear optimization is a key component of many image registration algorithms. Improving registration speed is almost always desirable. One way to do this is to accelerate the optimization cost function using a parallel implementation. The purpose of this document is to provide a tutorial on how to combine the CUDA GPU computing framework with standard nonlinear optimization libraries (VNL) using CMake. The provided code can be used as a starting template for programmers looking for a relatively painless introduction to CUDA-accelerated medical image registration and other nonlinear optimization problems.

A Tutorial on Combining Nonlinear Optimization with CUDA

Anisotropic Non-Linear Diffusion is a powerful image processing technique, which allows to simultaneously remove the noise and enhance sharp features in two or three dimensional images. Anisotropic Diffusion is understood here in the sense of Weickert, meaning that diffusion tensors are anisotropic and reflect the local orientation of image features. This is in contrast with the non-linear diffusion filter of Perona and Malik, which only involves scalar diffusion coefficients, in other words isotropic diffusion tensors. 
In this paper, we present an anisotropic non-linear diffusion technique we implemented in ITK. This technique is based on a recent adaptive scheme making the diffusion stable and requiring limited numerical resources. 