Ali Kamen

A survey of medical image registration on graphics hardware

The rapidly increasing performance of graphics processors, improving programming support and excellent performance-price ratio make graphics processing units (GPUs) a good option for a variety of computationally intensive tasks. Within this survey, we give an overview of GPU accelerated image registration. We address both, GPU experienced readers with an interest in accelerated image registration, as well as registration experts who are interested in using GPUs. We survey programming models and interfaces and analyze different approaches to programming on the GPU. We furthermore discuss the inherent advantages and challenges of current hardware architectures, which leads to a description of the details of the important building blocks for successful implementations.

Linear intensity-based image registration by Markov random fields and discrete optimization

We propose a framework for intensity-based registration of images by linear transformations, based on a discrete Markov random field (MRF) formulation. Here, the challenge arises from the fact that optimizing the energy associated with this problem requires a high-order MRF model. Currently, methods for optimizing such high-order models are less general, easy to use, and efficient, than methods for the popular second-order models. Therefore, we propose an approximation to the original energy by an MRF with tractable second-order terms. The approximation at a certain point p in the parameter space is the normalized sum of evaluations of the original energy at projections of p to two-dimensional subspaces. We demonstrate the quality of the proposed approximation by computing the correlation with the original energy, and show that registration can be performed by discrete optimization of the approximated energy in an iteration loop. A search space refinement strategy is employed over iterations to achieve sub-pixel accuracy, while keeping the number of labels small for efficiency. The proposed framework can encode any similarity measure is robust to the settings of the internal parameters, and allows an intuitive control of the parameter ranges. We demonstrate the applicability of the framework by intensity-based registration, and 2D-3D registration of medical images. The evaluation is performed by random studies and real registration tasks. The tests indicate increased robustness and precision compared to corresponding standard optimization of the original energy, and demonstrate robustness to noise. Finally, the proposed framework allows the transfer of advances in MRF optimization to linear registration problems.

Computer-aided gleason grading of prostate cancer histopathological images using texton forests

The Gleason score is the single most important prognostic indicator for prostate cancer candidates and plays a significant role in treatment planning. Histopathological imaging of prostate tissue samples provides the gold standard for obtaining the Gleason score, but the manual assignment of Gleason grades is a labor-intensive and error-prone process. We have developed a texture classification system for automatic and reproducible Gleason grading. Our system characterizes the texture in images belonging to a tumor grade by clustering extracted filter responses at each pixel into textons (basic texture elements). We have used random forests to cluster the filter responses into textons followed by the spatial pyramid match kernel in conjunction with an SVM classifier. We have demonstrated the efficacy of our system in distinguishing between Gleason grades 3 and 4.

A patient-specific reduced-order model for coronary circulation

We introduce a patient-specific model for coronary circulation, by combining anatomical, hemodynamic and functional information from medical images and other clinical observations. The main components of the coupled model are: a lumped heart model, a reduced-order model for hemodynamics in the arterial vessel tree (both healthy and stenosed), and a physiological model for the microvascular bed. The anatomy of the vessel tree is extracted from Coronary Computed Tomography Angiography (CTA) images, followed by an estimation of the impedance of the distal microvascular network. For the blood flow simulations, three states are modeled: rest, drug-induced hyperemia and intense exercise. The results show an excellent agreement with the literature and provide a model for virtual assessment of the flow and underlying functional measures in healthy and stenosed coronary arteries.

A framework for personalization of coronary flow computations during rest and hyperemia

We introduce a Computational Fluid Dynamics (CFD) based method for performing patient-specific coronary hemodynamic computations under two conditions: at rest and during drug-induced hyperemia. The proposed method is based on a novel estimation procedure for determining the boundary conditions from non-invasively acquired patient data at rest. A multi-variable feedback control framework ensures that the computed mean arterial pressure and the flow distribution matches the estimated values for an individual patient during the rest state. The boundary conditions at hyperemia are derived from the respective rest-state values via a transfer function that models the vasodilation phenomenon. Simulations are performed on a coronary tree where a 65% diameter stenosis is introduced in the left anterior descending (LAD) artery, with the boundary conditions estimated using the proposed method. The results demonstrate that the estimation of the hyperemic resistances is crucial in order to obtain accurate values for pressure and flow rates. Results from an exhaustive sensitivity analysis have been presented for analyzing the variability of trans-stenotic pressure drop and Fractional Flow Reserve (FFR) values with respect to various measurements and assumptions.

Automated malignancy detection in breast histopathological images

Detection of malignancy from histopathological images of breast cancer is a labor-intensive and error-prone process. To streamline this process, we present an efficient Computer Aided Diagnostic system that can differentiate between cancerous and non-cancerous H&E (hemotoxylin&eosin) biopsy samples. Our system uses novel textural, topological and morphometric features taking advantage of the special patterns of the nuclei cells in breast cancer histopathological images. We use a Support Vector Machine classifier on these features to diagnose malignancy. In conjunction with the maximum relevance - minimum redundancy feature selection technique, we obtain high sensitivity and specificity. We have also investigated the effect of image compression on classification performance.

An efficient graph-based deformable 2D/3D registration algorithm with applications for abdominal aortic aneurysm interventions

2D/3D registration is in general a challenging task due to its ill-posed nature. It becomes even more difficult when deformation between the 3D volume and 2D images needs to be recovered. This paper presents an automatic, accurate and efficient deformable 2D/3D registration method that is formulated on a 3D graph and applied for abdominal aortic aneurysm (AAA) interventions. The proposed method takes the 3D graph generated from a segmentation of the CT volume and the 2D distance map calculated from the 2D X-ray image as the input. The similarity measure consists of a difference measure, a length preservation term and a smoothness regularization term, all of which are defined and efficiently calculated on the graph. A hierarchical registration scheme is further designed specific to the anatomy of abdominal aorta and typical deformations observed during AAA cases. The method was validated using both phantom and clinical datasets, and achieved an average error of > 1mm within 0.1s. The proposed method is of general form and has the potential to be applied for a wide range of applications requiring efficient 2D/3D registration of vascular structures.

Network cycle features: Application to computer-aided Gleason grading of prostate cancer histopathological images

Features extracted from cell networks have become popular tools in histological image analysis. However, existing features do not take sufficient advantage of the cycle structure present within the cell networks. We introduce a new class of network cycle features that take advantage of such structures. We demonstrate the utility of these features for automated prostate cancer scoring using histological images. Prostate cancer is commonly scored by pathologists using the Gleason grading system and our automated system based upon network cycle features serves an important need in making this process less labor-intensive and more reproducible. Our system first extracts the cells from the histological images, computes networks from the cell locations and then computes features based upon statistics for the different cycles present in these networks. Using an SVM (Support Vector Machine) classifier on these features, we demonstrate the efficacy of our system in distinguishing between grade 3 and grade 4 prostate tumors. We also show the superiority of our approach over previously developed systems for this problem based upon texture features, fractal features and alternative network features.

Comparison of Fractional Flow Reserve Based on Computational Fluid Dynamics Modeling Using Coronary Angiographic Vessel Morphology Versus Invasively Measured Fractional Flow Reserve

Invasive fractional flow reserve (FFRinvasive), although gold standard to identify hemodynamically relevant coronary stenoses, is time consuming and potentially associated with complications. We developed and evaluated a new approach to determine lesion-specific FFR on the basis of coronary anatomy as visualized by invasive coronary angiography (FFRangio): 100 coronary lesions (50% to 90% diameter stenosis) in 73 patients (48 men, 25 women; mean age 67 ± 9 years) were studied. On the basis of coronary angiograms acquired at rest from 2 views at angulations at least 30° apart, a PC-based computational fluid dynamics modeling software used personalized boundary conditions determined from 3-dimensional reconstructed angiography, heart rate, and blood pressure to derive FFRangio. The results were compared with FFRinvasive. Interobserver variability was determined in a subset of 25 narrowings. Twenty-nine of 100 coronary lesions were hemodynamically significant (FFRinvasive ≤ 0.80). FFRangio identified these with an accuracy of 90%, sensitivity of 79%, specificity of 94%, positive predictive value of 85%, and negative predictive value of 92%. The area under the receiver operating characteristic curve was 0.93. Correlation between FFRinvasive (mean: 0.84 ± 0.11) and FFRangio (mean: 0.85 ± 0.12) was r = 0.85. Interobserver variability of FFRangio was low, with a correlation of r = 0.88. In conclusion, estimation of coronary FFR with PC-based computational fluid dynamics modeling on the basis of lesion morphology as determined by invasive angiography is possible with high diagnostic accuracy compared to invasive measurements.

Efficient Lattice Boltzmann Solver for Patient-Specific Radiofrequency Ablation of Hepatic Tumors

Radiofrequency ablation (RFA) is an established treatment for liver cancer when resection is not possible. Yet, its optimal delivery is challenged by the presence of large blood vessels and the time-varying thermal conductivity of biological tissue. Incomplete treatment and an increased risk of recurrence are therefore common. A tool that would enable the accurate planning of RFA is hence necessary. This manuscript describes a new method to compute the extent of ablation required based on the Lattice Boltzmann Method (LBM) and patient-specific, pre-operative images. A detailed anatomical model of the liver is obtained from volumetric images. Then a computational model of heat diffusion, cellular necrosis, and blood flow through the vessels and liver is employed to compute the extent of ablated tissue given the probe location, ablation duration and biological parameters. The model was verified against an analytical solution, showing good fidelity. We also evaluated the predictive power of the proposed framework on ten patients who underwent RFA, for whom pre- and post-operative images were available. Comparisons between the computed ablation extent and ground truth, as observed in postoperative images, were promising (DICE index: 42%, sensitivity: 67%, positive predictive value: 38%). The importance of considering liver perfusion while simulating electrical-heating ablation was also highlighted. Implemented on graphics processing units (GPU), our method simulates 1 minute of ablation in 1.14 minutes, allowing near real-time computation.