Search

Current clinical practice relies on clinical history to determine the time since stroke (TSS) onset. Imaging-based determination of acute stroke onset time could provide critical information to clinicians in deciding stroke treatment options, such as thrombolysis. The patients with unknown or unwitnessed TSS are usually excluded from thrombolysis, even if their symptoms began within the therapeutic window. In this paper, we demonstrate a machine learning approach for TSS classification using routinely acquired imaging sequences. We develop imaging features from the magnetic resonance (MR) images and train machine learning models to classify the TSS. We also propose a deep-learning model to extract hidden representations for the MR perfusion-weighted images and demonstrate classification improvement by incorporating these additional deep features. The cross-validation results show that our best classifier achieved an area under the curve of 0.765, with a sensitivity of 0.788 and a negative predictive value of 0.609, outperforming existing methods. We show that the features generated by our deep-learning algorithm correlate with the MR imaging features, and validate the robustness of the model on imaging parameter variations (e.g., year of imaging). This paper advances magnetic resonance imaging analysis one-step-closer to an operational decision support tool for stroke treatment guidance.

Image reconstruction of low-count positron emission tomography (PET) data is challenging. Kernel methods address the challenge by incorporating image prior information in the forward model of iterative PET image reconstruction. The kernelized expectation-maximization (KEM) algorithm has been developed and demonstrated to be effective and easy to implement. A common approach for a further improvement of the kernel method would be adding an explicit regularization, which however leads to a complex optimization problem. In this paper, we propose an implicit regularization for the kernel method by using a deep coefficient prior, which represents the kernel coefficient image in the PET forward model using a convolutional neural-network. To solve the maximum-likelihood neural network-based reconstruction problem, we apply the principle of optimization transfer to derive a neural KEM algorithm. Each iteration of the algorithm consists of two separate steps: a KEM step for image update from the projection data and a deep-learning step in the image domain for updating the kernel coefficient image using the neural network. This optimization algorithm is guaranteed to monotonically increase the data likelihood. The results from computer simulations and real patient data have demonstrated that the neural KEM can outperform existing KEM and deep image prior methods.

Prostate cancer is the second leading cause of cancer death among men in the United States. The diagnosis of prostate MRI often relies on accurate prostate zonal segmentation. However, state-of-the-art automatic segmentation methods often fail to produce well-contained volumetric segmentation of the prostate zones since certain slices of prostate MRI, such as base and apex slices, are harder to segment than other slices. This difficulty can be overcome by leveraging important multi-scale image-based information from adjacent slices, but current methods do not fully learn and exploit such cross-slice information. In this paper, we propose a novel cross-slice attention mechanism, which we use in a Transformer module to systematically learn cross-slice information at multiple scales. The module can be utilized in any existing deep-learning-based segmentation framework with skip connections. Experiments show that our cross-slice attention is able to capture cross-slice information significant for prostate zonal segmentation in order to improve the performance of current state-of-the-art methods. Cross-slice attention improves segmentation accuracy in the peripheral zones, such that segmentation results are consistent across all the prostate slices (apex, mid-gland, and base). The code for the proposed model is available at https://bit.ly/CAT-Net.

Image reconstruction for positron emission tomography (PET) is challenging because of the ill-conditioned tomographic problem and low counting statistics. Kernel methods address this challenge by using kernel representation to incorporate image prior information in the forward model of iterative PET image reconstruction. Existing kernel methods construct the kernels commonly using an empirical process, which may lead to unsatisfactory performance. In this paper, we describe the equivalence between the kernel representation and a trainable neural network model. A deep kernel method is then proposed by exploiting a deep neural network to enable automated learning of an improved kernel model and is directly applicable to single subjects in dynamic PET. The training process utilizes available image prior data to form a set of robust kernels in an optimized way rather than empirically. The results from computer simulations and a real patient dataset demonstrate that the proposed deep kernel method can outperform the existing kernel method and neural network method for dynamic PET image reconstruction.

Accurately segmenting teeth and identifying the corresponding anatomical landmarks on dental mesh models are essential in computer-aided orthodontic treatment. Manually performing these two tasks is time-consuming, tedious, and, more importantly, highly dependent on orthodontists experiences due to the abnormality and large-scale variance of patients teeth. Some machine learning-based methods have been designed and applied in the orthodontic field to automatically segment dental meshes (e.g., intraoral scans). In contrast, the number of studies on tooth landmark localization is still limited. This paper proposes a two-stage framework based on mesh deep learning (called TS-MDL) for joint tooth labeling and landmark identification on raw intraoral scans. Our TS-MDL first adopts an end-to-end iMeshSegNet method (i.e., a variant of the existing MeshSegNet with both improved accuracy and efficiency) to label each tooth on the downsampled scan. Guided by the segmentation outputs, our TS-MDL further selects each tooths region of interest (ROI) on the original mesh to construct a light-weight variant of the pioneering PointNet (i.e., PointNet-Reg) for regressing the corresponding landmark heatmaps. Our TS-MDL was evaluated on a real-clinical dataset, showing promising segmentation and localization performance. Specifically, iMeshSegNet in the first stage of TS-MDL reached an averaged Dice similarity coefficient (DSC) at 0.964±0.054 , significantly outperforming the original MeshSegNet. In the second stage, PointNet-Reg achieved a mean absolute error (MAE) of 0.597±0.761 mm in distances between the prediction and ground truth for 66 landmarks, which is superior compared with other networks for landmark detection. All these results suggest the potential usage of our TS-MDL in orthodontics.

Tensor fields are useful for modeling the structure of biological tissues. The challenge to measure tensor fields involves acquiring sufficient data of scalar measurements that are physically achievable and reconstructing tensors from as few projections as possible for efficient applications in medical imaging. In this paper, we present a filtered back-projection algorithm for the reconstruction of a symmetric second-rank tensor field from directional X-ray projections about three axes. The tensor field is decomposed into a solenoidal and irrotational component, each of three unknowns. Using the Fourier projection theorem, a filtered back-projection algorithm is derived to reconstruct the solenoidal and irrotational components from projections acquired around three axes. A simple illustrative phantom consisting of two spherical shells and a 3D digital cardiac diffusion image obtained from diffusion tensor MRI of an excised human heart are used to simulate directional X-ray projections. The simulations validate the mathematical derivations and demonstrate reasonable noise properties of the algorithm. The decomposition of the tensor field into solenoidal and irrotational components provides insight into the development of algorithms for reconstructing tensor fields with sufficient samples in terms of the type of directional projections and the necessary orbits for the acquisition of the projections of the tensor field.

Deep neural networks, in particular convolutional networks, have rapidly become a popular choice for analyzing histopathology images. However, training these models relies heavily on a large number of samples manually annotated by experts, which is cumbersome and expensive. In addition, it is difficult to obtain a perfect set of labels due to the variability between expert annotations. This paper presents a novel active learning (AL) framework for histopathology image analysis, named PathAL. To reduce the required number of expert annotations, PathAL selects two groups of unlabeled data in each training iteration: one "informative" sample that requires additional expert annotation, and one "confident predictive" sample that is automatically added to the training set using the model's pseudo-labels. To reduce the impact of the noisy-labeled samples in the training set, PathAL systematically identifies noisy samples and excludes them to improve the generalization of the model. Our model advances the existing AL method for medical image analysis in two ways. First, we present a selection strategy to improve classification performance with fewer manual annotations. Unlike traditional methods focusing only on finding the most uncertain samples with low prediction confidence, we discover a large number of high confidence samples from the unlabeled set and automatically add them for training with assigned pseudo-labels. Second, we design a method to distinguish between noisy samples and hard samples using a heuristic approach. We exclude the noisy samples while preserving the hard samples to improve model performance. Extensive experiments demonstrate that our proposed PathAL framework achieves promising results on a prostate cancer Gleason grading task, obtaining similar performance with 40% fewer annotations compared to the fully supervised learning scenario. An ablation study is provided to analyze the effectiveness of each component in PathAL, and a pathologist reader study is conducted to validate our proposed algorithm.

Recently, deep neural networks have been widely and successfully applied in computer vision tasks and have attracted growing interest in medical imaging. One barrier for the application of deep neural networks to medical imaging is the need for large amounts of prior training pairs, which is not always feasible in clinical practice. This is especially true for medical image reconstruction problems, where raw data are needed. Inspired by the deep image prior framework, in this paper, we proposed a personalized network training method where no prior training pairs are needed, but only the patient's own prior information. The network is updated during the iterative reconstruction process using the patient-specific prior information and measured data. We formulated the maximum-likelihood estimation as a constrained optimization problem and solved it using the alternating direction method of multipliers algorithm. Magnetic resonance imaging guided positron emission tomography reconstruction was employed as an example to demonstrate the effectiveness of the proposed framework. Quantification results based on simulation and real data show that the proposed reconstruction framework can outperform Gaussian post-smoothing and anatomically guided reconstructions using the kernel method or the neural-network penalty.

Current clinical dynamic PET has an effective temporal resolution of 5-10 seconds, which can be adequate for traditional compartmental modeling but is inadequate for exploiting the benefit of more advanced tracer kinetic modeling for characterization of diseases (e.g., cancer and heart disease). There is a need to improve dynamic PET to allow fine temporal sampling of 1-2 seconds. However, the reconstruction of these short-time frames from tomographic data is extremely challenging as the count level of each frame is very low and high noise presents in both spatial and temporal domains. Previously, the kernel framework has been developed and demonstrated as a statistically efficient approach to utilizing image prior for low-count PET image reconstruction. Nevertheless, the existing kernel methods mainly explore spatial correlations in the data and only have a limited ability in suppressing temporal noise. In this paper, we propose a new kernel method which extends the previous spatial kernel method to the general spatiotemporal domain. The new kernelized model encodes both spatial and temporal correlations obtained from image prior information and are incorporated into the PET forward projection model to improve themaximumlikelihood(ML) image reconstruction. Computer simulations and an application to real patient scan have shown that the proposed approach can achieve effective noise reduction in both spatial and temporal domains and outperform the spatial kernel method and conventional ML reconstruction method for improving the high temporal-resolution dynamic PET imaging.

PET image reconstruction is challenging due to the ill-poseness of the inverse problem and limited number of detected photons. Recently, the deep neural networks have been widely and successfully used in computer vision tasks and attracted growing interests in medical imaging. In this paper, we trained a deep residual convolutional neural network to improve PET image quality by using the existing inter-patient information. An innovative feature of the proposed method is that we embed the neural network in the iterative reconstruction framework for image representation, rather than using it as a post-processing tool. We formulate the objective function as a constrained optimization problem and solve it using the alternating direction method of multipliers algorithm. Both simulation data and hybrid real data are used to evaluate the proposed method. Quantification results show that our proposed iterative neural network method can outperform the neural network denoising and conventional penalized maximum likelihood methods.

In this paper, we study a central problem in multimodal neuroimaging analysis, i.e., identification of significantly correlated brain regions between multiple imaging modalities. We propose a spatially varying correlation model and the associated inference procedure, which improves substantially over the common alternative solutions of voxel-wise and region-wise analysis. Compared with voxel-wise analysis, our method aggregates voxels with similar correlations into regions, takes into account spatial continuity of correlations at nearby voxels, and enjoys a much higher detection power. Compared with region-wise analysis, our method does not rely on any pre-specified brain region map, but instead finds homogenous correlation regions adaptively given the data. We applied our method to a multimodal positron emission tomography study, and found brain regions with significant correlation between tau and glucose metabolism that voxel-wise or region-wise analysis failed to identify. Our findings conform and lend additional support to prior hypotheses about how the two pathological proteins of Alzheimers disease, tau and amyloid, interact with glucose metabolism in the aging human brain.

We present a method of generating spatial maps of kinetic parameters from dynamic sequences of images collected in hyperpolarized carbon-13 magnetic resonance imaging (MRI) experiments. The technique exploits spatial correlations in the dynamic traces via regularization in the space of parameter maps. Similar techniques have proven successful in other dynamic imaging problems, such as dynamic contrast enhanced MRI. In this paper, we apply these techniques for the first time to hyperpolarized MRI problems, which are particularly challenging due to limited signal-to-noise ratio (SNR). We formulate the reconstruction as an optimization problem and present an efficient iterative algorithm for solving it based on the alternation direction method of multipliers. We demonstrate that this technique improves the qualitative appearance of parameter maps estimated from low SNR dynamic image sequences, first in simulation then on a number of data sets collected in vivo. The improvement this method provides is particularly pronounced at low SNR levels.

Tomographic image reconstruction for low-dose computed tomography (CT) is increasingly challenging as dose continues to reduce in clinical applications. Pre-log domain methods and post-log domain methods have been proposed individually and each method has its own disadvantage. While having the potential to improve image quality for low-dose data by using an accurate imaging model, pre-log domain methods suffer slow convergence in practice due to the nonlinear transformation from the image to measurements. In contrast, post-log domain methods have fast convergence speed but the resulting image quality is suboptimal for low dose CT data because the log transformation is extremely unreliable for low-count measurements and undefined for negative values. This paper proposes a hybrid method that integrates the pre-log model and post-log model together to overcome the disadvantages of individual pre-log and post-log methods. We divide a set of CT data into high-count and low-count regions. The post-log weighted least squares model is used for measurements in the high-count region and the pre-log shifted Poisson model for measurements in the low-count region. The hybrid likelihood function can be optimized using an existing iterative algorithm. Computer simulations and phantom experiments show that the proposed hybrid method can achieve faster early convergence than the pre-log shifted Poisson likelihood method and better signal-to-noise performance than the post-log weighted least squares method.

Iterative image reconstruction for positron emission tomography can improve image quality by using spatial regularization. The most commonly used quadratic penalty often oversmoothes sharp edges and fine features in reconstructed images, while nonquadratic penalties can preserve edges and achieve higher contrast recovery. Existing optimization algorithms such as the expectation maximization (EM) and preconditioned conjugate gradient (PCG) algorithms work well for the quadratic penalty, but are less efficient for high-curvature or nonsmooth edge-preserving regularizations. This paper proposes a new algorithm to accelerate edge-preserving image reconstruction by using two strategies: trust surrogate and optimization transfer descent. Trust surrogate approximates the original penalty by a smoother function at each iteration, but guarantees the algorithm to descend monotonically; Optimization transfer descent accelerates a conventional optimization transfer algorithm by using conjugate gradient and line search. Results of computer simulations and real 3-D data show that the proposed algorithm converges much faster than the conventional EM and PCG for smooth edge-preserving regularization and can also be more efficient than the current state-of-art algorithms for the nonsmooth l1 regularization.

Image reconstruction from low-count positron emission tomography (PET) projection data is challenging because the inverse problem is ill-posed. Prior information can be used to improve image quality. Inspired by the kernel methods in machine learning, this paper proposes a kernel based method that models PET image intensity in each pixel as a function of a set of features obtained from prior information. The kernel-based image model is incorporated into the forward model of PET projection data and the coefficients can be readily estimated by the maximum likelihood (ML) or penalized likelihood image reconstruction. A kernelized expectation-maximization algorithm is presented to obtain the ML estimate. Computer simulations show that the proposed approach can achieve better bias versus variance trade-off and higher contrast recovery for dynamic PET image reconstruction than the conventional maximum likelihood method with and without post-reconstruction denoising. Compared with other regularization-based methods, the kernel method is easier to implement and provides better image quality for low-count data. Application of the proposed kernel method to a 4-D dynamic PET patient dataset showed promising results.

Fluorescence lifetime imaging (FLIM) has demonstrated potential for robust assessment of atherosclerotic plaques biochemical composition and for complementing conventional intravascular ultrasound (IVUS), which provides information on plaque morphology. The success of such a bi-modal imaging modality depends on accurate segmentation of the IVUS images and proper angular registration between these two modalities. This paper reports a novel IVUS segmentation methodology addressing this issue. The image preprocessing consisted of denoising, using the Wiener filter, followed by image smoothing, implemented through the application of the alternating sequential filter on the edge separability metric images. Extraction of the lumen/intima and media/adventitia boundaries was achieved by tracing the gray-scale peaks over the A-lines of the IVUS preprocessed images. Cubic spline interpolation, in both cross-sectional and longitudinal directions, ensured boundary smoothness and continuity. The detection of the guide-wire artifact in both modalities is used for angular registration. Intraluminal studies were conducted in 13 ex vivo segments of human coronaries. The IVUS segmentation accuracy was assessed against independent manual tracings, providing 91.82% sensitivity and 97.55% specificity. The proposed methodology makes the bi-modal FLIM and IVUS approach feasible for comprehensive intravascular diagnosis by providing co-registered biochemical and morphological information of atherosclerotic plaques.

In this paper, we present a novel approach for the intrinsic mapping of anatomical surfaces and its application in brain mapping research. Using the Laplace-Beltrami eigen-system, we represent each surface with an isometry invariant embedding in a high dimensional space. The key idea in our system is that we realize surface deformation in the embedding space via the iterative optimization of a conformal metric without explicitly perturbing the surface or its embedding. By minimizing a distance measure in the embedding space with metric optimization, our method generates a conformal map directly between surfaces with highly uniform metric distortion and the ability of aligning salient geometric features. Besides pairwise surface maps, we also extend the metric optimization approach for group-wise atlas construction and multi-atlas cortical label fusion. In experimental results, we demonstrate the robustness and generality of our method by applying it to map both cortical and hippocampal surfaces in population studies. For cortical labeling, our method achieves excellent performance in a cross-validation experiment with 40 manually labeled surfaces, and successfully models localized brain development in a pediatric study of 80 subjects. For hippocampal mapping, our method produces much more significant results than two popular tools on a multiple sclerosis study of 109 subjects.

In this paper, we present a novel approach for the intrinsic mapping of anatomical surfaces and its application in brain mapping research. Using the Laplace-Beltrami eigen-system, we represent each surface with an isometry invariant embedding in a high dimensional space. The key idea in our system is that we realize surface deformation in the embedding space via the iterative optimization of a conformal metric without explicitly perturbing the surface or its embedding. By minimizing a distance measure in the embedding space with metric optimization, our method generates a conformal map directly between surfaces with highly uniform metric distortion and the ability of aligning salient geometric features. Besides pairwise surface maps, we also extend the metric optimization approach for group-wise atlas construction and multi-atlas cortical label fusion. In experimental results, we demonstrate the robustness and generality of our method by applying it to map both cortical and hippocampal surfaces in population studies. For cortical labeling, our method achieves excellent performance in a cross-validation experiment with 40 manually labeled surfaces, and successfully models localized brain development in a pediatric study of 80 subjects. For hippocampal mapping, our method produces much more significant results than two popular tools on a multiple sclerosis study of 109 subjects.

Proton radiography has applications in patient alignment and verification procedures for proton beam radiation therapy. In this paper, we report an experiment which used 200 MeV protons to generate proton energy-loss and scattering radiographs of a hand phantom. The experiment used the first-generation proton computed tomography (CT) scanner prototype, which was installed on the research beam line of the clinical proton synchrotron at Loma Linda University Medical Center. It was found that while both radiographs displayed anatomical details of the hand phantom, the energy-loss radiograph had a noticeably higher resolution. Nonetheless, scattering radiography may yield more contrast between soft and bone tissue than energy-loss radiography, however, this requires further study. This study contributes to the optimization of the performance of the next-generation of clinical proton CT scanners. Furthermore, it demonstrates the potential of proton imaging (proton radiography and CT), which is now within reach of becoming available as a new, potentially low-dose medical imaging modality.

Proton radiography has applications in patient alignment and verification procedures for proton beam radiation therapy. In this paper, we report an experiment which used 200 MeV protons to generate proton energy-loss and scattering radiographs of a hand phantom. The experiment used the first-generation proton computed tomography (CT) scanner prototype, which was installed on the research beam line of the clinical proton synchrotron at Loma Linda University Medical Center. It was found that while both radiographs displayed anatomical details of the hand phantom, the energy-loss radiograph had a noticeably higher resolution. Nonetheless, scattering radiography may yield more contrast between soft and bone tissue than energy-loss radiography, however, this requires further study. This study contributes to the optimization of the performance of the next-generation of clinical proton CT scanners. Furthermore, it demonstrates the potential of proton imaging (proton radiography and CT), which is now within reach of becoming available as a new, potentially low-dose medical imaging modality.

Currently, there is a lack of computational methods for the evaluation of mild traumatic brain injury (mTBI) from magnetic resonance imaging (MRI). Further, the development of automated analyses has been hindered by the subtle nature of mTBI abnormalities, which appear as low contrast MR regions. This paper proposes an approach that is able to detect mTBI lesions by combining both the high-level context and low-level visual information. The contextual model estimates the progression of the disease using subject information, such as the time since injury and the knowledge about the location of mTBI. The visual model utilizes texture features in MRI along with a probabilistic support vector machine to maximize the discrimination in unimodal MR images. These two models are fused to obtain a final estimate of the locations of the mTBI lesion. The models are tested using a novel rodent model of repeated mTBI dataset. The experimental results demonstrate that the fusion of both contextual and visual textural features outperforms other state-of-the-art approaches. Clinically, our approach has the potential to benefit both clinicians by speeding diagnosis and patients by improving clinical care.

Currently, there is a lack of computational methods for the evaluation of mild traumatic brain injury (mTBI) from magnetic resonance imaging (MRI). Further, the development of automated analyses has been hindered by the subtle nature of mTBI abnormalities, which appear as low contrast MR regions. This paper proposes an approach that is able to detect mTBI lesions by combining both the high-level context and low-level visual information. The contextual model estimates the progression of the disease using subject information, such as the time since injury and the knowledge about the location of mTBI. The visual model utilizes texture features in MRI along with a probabilistic support vector machine to maximize the discrimination in unimodal MR images. These two models are fused to obtain a final estimate of the locations of the mTBI lesion. The models are tested using a novel rodent model of repeated mTBI dataset. The experimental results demonstrate that the fusion of both contextual and visual textural features outperforms other state-of-the-art approaches. Clinically, our approach has the potential to benefit both clinicians by speeding diagnosis and patients by improving clinical care.

Parallel imaging is a powerful technique to speed up magnetic resonance (MR) image acquisition via multiple coils. Both the received signal of each coil and its sensitivity map, which describes its spatial response, are needed during reconstruction. Widely used schemes such as SENSE assume that sensitivity maps of the coils are noiseless while the only errors are in coil outputs. In practice, however, sensitivity maps are subject to a wide variety of errors. At first glance, sensitivity noise appears to result in an errors-in-variables problem of the kind that is typically solved using total least squares (TLSs). However, existing TLS algorithms are in general inappropriate for the specific type of block structure that arises in parallel imaging. In this paper, we take a maximum likelihood approach to the problem of parallel imaging in the presence of independent Gaussian sensitivity noise. This results in a quasi-quadratic objective function, which can be efficiently minimized. Experimental evidence suggests substantial gains over conventional SENSE, especially in nonideal imaging conditions like low signal-to-noise ratio (SNR), high g-factors and large acceleration, using sensitivity maps suffering from misalignment, ringing, and random noise.