Unsupervised brain lesion segmentation, focusing on learning normative distributions from images of healthy subjects, are less dependent on lesion-labeled data, thus exhibiting better generalization capabilities. A fundamental challenge in learning normative distributions of images lies in the high dimensionality if image pixels are treated as correlated random variables to capture spatial dependence. In this study, we proposed a subspace-based deep generative model to learn the posterior normal distributions. Specifically, we used probabilistic subspace models to capture spatial-intensity distributions and spatial-structure distributions of brain images from healthy subjects. These models captured prior spatial-intensity and spatial-structure variations effectively by treating the subspace coefficients as random variables with basis functions being the eigen-images and eigen-density functions learned from the training data. These prior distributions were then converted to posterior distributions, including both the posterior normal and posterior lesion distributions for a given image using the subspace-based generative model and subspace-assisted Bayesian analysis, respectively. Finally, an unsupervised fusion classifier was used to combine the posterior and likelihood features for lesion segmentation. The proposed method has been evaluated on simulated and real lesion data, including tumor, multiple sclerosis, and stroke, demonstrating superior segmentation accuracy and robustness over the state-of-the-art methods. Our proposed method holds promise for enhancing unsupervised brain lesion delineation in clinical applications.

In volume-to-volume translations in medical images, existing models often struggle to capture the inherent volumetric distribution using 3D voxel-space representations, due to high computational dataset demands. We present Score-Fusion, a novel volumetric translation model that effectively learns 3D representations by ensembling perpendicularly trained 2D diffusion models in score function space. By carefully initializing our model to start with an average of 2D models as in existing models, we reduce 3D training to a fine-tuning process, mitigating computational and data demands. Furthermore, we explicitly design the 3D model’s hierarchical layers to learn ensembles of 2D features, further enhancing efficiency and performance. Moreover, Score-Fusion naturally extends to multi-modality settings by fusing diffusion models conditioned on different inputs for flexible, accurate integration. We demonstrate that 3D representation is essential for better performance in downstream recognition tasks, such as tumor segmentation, where most segmentation models are based on 3D representation. Extensive experiments demonstrate that Score-Fusion achieves superior accuracy and volumetric fidelity in 3D medical image super-resolution and modality translation. Additionally, we extend Score-Fusion to video super-resolution by integrating 2D diffusion models on time-space slices with a spatial-temporal video diffusion backbone, highlighting its potential for general-purpose volume translation and providing broader insight into learning-based approaches for score function fusion.

Multimodal image translation has found useful applications in solving several medical imaging problems. In this paper, we presented a systematic analysis of multimodal images and machine learning-based image translation from an information-theoretic perspective. Specifically, we analyzed the amount of mutual information that exists in some commonly used multimodal images. This analysis revealed varying structural correlation across modalities and tissue-dependence of mutual information. We also analyzed the amount of information transferred and gained in multimodal image translation and provided an upper bound on the information gain. Information-theoretic measures were also proposed to assess the effectiveness of an image translator, and the uncertainty associated with image translation. Numerical results were presented to demonstrate the information gain in practical multimodal image translation, and to validate the proposed upper bound on information gain and the translation error predictor. Finally, several potential applications of our analysis results were discussed, including the image denoising and reconstruction using side information generated by image translation. The findings from this study may prove useful for guiding the further development and application of multimodal image translation.

Image reconstruction from limited and/or sparse data is known to be an ill-posed problem and a priori information/constraints have played an important role in solving the problem. Early constrained image reconstruction methods utilize image priors based on general image properties such as sparsity, low-rank structures, spatial support bound, etc. Recent deep learning-based reconstruction methods promise to produce even higher quality reconstructions by utilizing more specific image priors learned from training data. However, learning high-dimensional image priors requires huge amounts of training data that are currently not available in medical imaging applications. As a result, deep learning-based reconstructions often suffer from two known practical issues: a) sensitivity to data perturbations (e.g., changes in data sampling scheme), and b) limited generalization capability (e.g., biased reconstruction of lesions). This paper proposes a new method to address these issues. The proposed method synergistically integrates model-based and data-driven learning in three key components. The first component uses the linear vector space framework to capture global dependence of image features; the second exploits a deep network to learn the mapping from a linear vector space to a nonlinear manifold; the third is an unrolling-based deep network that captures local residual features with the aid of a sparsity model. The proposed method has been evaluated with magnetic resonance imaging data, demonstrating improved reconstruction in the presence of data perturbation and/or novel image features. The method may enhance the practical utility of deep learning-based image reconstruction.

Deuterium magnetic resonance spectroscopic imaging (DMRSI) has recently been recognized as a potentially powerful tool for noninvasive imaging of brain energy metabolism and tumor. However, the low sensitivity of DMRSI has significantly limited its utility for both research and clinical applications. This work presents a novel machine learning-based method to address this limitation. The proposed method synergistically integrates physics-based subspace modeling and data-driven deep learning for effective denoising, making high-resolution dynamic DMRSI possible. Specifically, a novel subspace model was used to represent the dynamic DMRSI signals; deep neural networks were trained to capture the low-dimensional manifolds of the spectral and temporal distributions of practical dynamic DMRSI data. The learned subspace and manifold structures were integrated via a regularization formulation to remove measurement noise. Theoretical analysis, computer simulations, and in vivo experiments have been conducted to demonstrate the denoising efficacy of the proposed method which enabled high-resolution imaging capability. The translational potential was demonstrated in tumor-bearing rats, where the Warburg effect associated with cancer metabolism and tumor heterogeneity were successfully captured. The new method may not only provide an effective tool to enhance the sensitivity of DMRSI for basic research and clinical applications but also provide a framework for denoising other spatiospectral data.

High-dimensional MR imaging often requires long data acquisition time, thereby limiting its practical applications. This paper presents a low-rank tensor based method for accelerated high-dimensional MR imaging using sparse sampling. This method represents high-dimensional images as low-rank tensors (or partially separable functions) and uses this mathematical structure for sparse sampling of the data space and for image reconstruction from highly undersampled data. More specifically, the proposed method acquires two datasets with complementary sampling patterns, one for subspace estimation and the other for image reconstruction; image reconstruction from highly undersampled data is accomplished by fitting the measured data with a sparsity constraint on the core tensor and a group sparsity constraint on the spatial coefficients jointly using the alternating direction method of multipliers. The usefulness of the proposed method is demonstrated in MRI applications; it may also have applications beyond MRI.

Magnetic resonance parameter mapping (e.g., T1 mapping, T2 mapping, T*2 mapping) is a valuable tool for tissue characterization. However, its practical utility has been limited due to long data acquisition time. This paper addresses this problem with a new model-based parameter mapping method. The proposed method utilizes a formulation that integrates the explicit signal model with sparsity constraints on the model parameters, enabling direct estimation of the parameters of interest from highly undersampled, noisy k-space data. An efficient greedy-pursuit algorithm is described to solve the resulting constrained parameter estimation problem. Estimation-theoretic bounds are also derived to analyze the benefits of incorporating sparsity constraints and benchmark the performance of the proposed method. The theoretical properties and empirical performance of the proposed method are illustrated in a T2 mapping application example using computer simulations.

Complexity in the acoustics of nasal vowels has long been acknowledged but complexity in their articulation has received less attention. A growing body of research suggests that velopharyngeal (VP) opening is complemented by other articulatory gestures which may enhance or counteract the acoustic outcomes of VP opening. In this paper we consider the role of pharyngeal aperture and lingual position in producing the phonemic distinction between oral and nasal vowels in Northern Metropolitan French. The results of a real-time MRI study of one female speaker confirm earlier findings related to tongue height and retraction. The results also suggest a role for the lower pharynx in centralizing the F1 of nasal vowels. Consideration is also given to the effect of the lowered velum on the acoustic transfer function of the oral tract of nasal vowels. We conclude that these articulations enhance some of the wellknown acoustic consequences of VP coupling associated with the production of nasal vowels. This supports and extends the hypothesis that the acoustic characteristics of nasalization can be attained by a family of speech gestures that include, but are not limited to, the opening of the VP port.

Equation 4 of the above-named article [ibid., vol. 30, no. 4, pp. 893-903, Apr. 2011] is corrected herein.

Partial separability (PS) and sparsity have been previously used to enable reconstruction of dynamic images from undersampled (k,t)-space data. This paper presents a new method to use PS and sparsity constraints jointly for enhanced performance in this context. The proposed method combines the complementary advantages of PS and sparsity constraints using a unified formulation, achieving significantly better reconstruction performance than using either of these constraints individually. A globally convergent computational algorithm is described to efficiently solve the underlying optimization problem. Reconstruction results from simulated and in vivo cardiac MRI data are also shown to illustrate the performance of the proposed method.

Compressed sensing (CS) has the potential to reduce magnetic resonance (MR) data acquisition time. In order for CS-based imaging schemes to be effective, the signal of interest should be sparse or compressible in a known representation, and the measurement scheme should have good mathematical properties with respect to this representation. While MR images are often compressible, the second requirement is often only weakly satisfied with respect to commonly used Fourier encoding schemes. This paper investigates the use of random encoding for CS-MRI, in an effort to emulate the "universal" encoding schemes suggested by the theoretical CS literature. This random encoding is achieved experimentally with tailored spatially-selective radio-frequency (RF) pulses. Both simulation and experimental studies were conducted to investigate the imaging properties of this new scheme with respect to Fourier schemes. Results indicate that random encoding has the potential to outperform conventional encoding in certain scenarios. However, our study also indicates that random encoding fails to satisfy theoretical sufficient conditions for stable and accurate CS reconstruction in many scenarios of interest. Therefore, there is still no general theoretical performance guarantee for CS-MRI, with or without random encoding, and CS-based methods should be developed and validated carefully in the context of specific applications.

Model-based techniques have the potential to reduce the artifacts and improve resolution in magnetic resonance spectroscopic imaging, without sacrificing the signal-to-noise ratio. However, the current approaches have a few drawbacks that limit their performance in practical applications. Specifically, the classical schemes use less flexible image models that lead to model misfit, thus resulting in artifacts. Moreover, the performance of the current approaches is negatively affected by the magnetic field inhomogeneity and spatial mismatch between the anatomical references and spectroscopic imaging data. In this paper, we propose efficient solutions to overcome these problems. We introduce a more flexible image model that represents the signal as a linear combination of compartmental and local basis functions. The former set represents the signal variations within the compartments, while the latter captures the local perturbations resulting from lesions or segmentation errors. Since the combined set is redundant, we obtain the reconstructions using sparsity penalized optimization. To compensate for the artifacts resulting from field inhomogeneity, we estimate the field map using alternate scans and use it in the reconstruction. We model the spatial mismatch as an affine transformation, whose parameters are estimated from the spectroscopy data.

Phase unwrapping is an important problem in many magnetic resonance imaging applications, such as field mapping and flow imaging. The challenge in two-dimensional phase unwrapping lies in distinguishing jumps due to phase wrapping from those due to noise and/or abrupt variations in the actual function. This paper addresses this problem using a Markov random field to model the true phase function, whose parameters are determined by maximizing the a posteriori probability. To reduce the computational complexity of the optimization procedure, an efficient algorithm is also proposed for parameter estimation using a series of dynamic programming connected by the iterated conditional modes. The proposed method has been tested with both simulated and experimental data, yielding better results than some of the state-of-the-art method (e.g., the popular least-squares method) in handling noisy phase images with rapid phase variations.

This paper presents a novel approach for landmark-based shape deformation, in which fitting error and shape difference are formulated into a support vector machine (SVM) regression problem. To well describe nonrigid shape deformation, this paper measures the shape difference using a thin-plate spline model. The proposed approach is capable of preserving the topology of the template shape in the deformation. This property is achieved by inserting a set of additional points and imposing a set of linear equality and/or inequality constraints. The underlying optimization problem is solved using a quadratic programming algorithm. The proposed method has been tested using practical data in the context of shape-based image segmentation. Some relevant practical issues, such as missing detected landmarks and selection of the regularization parameter are also briefly discussed.

This paper presents an analysis of the mutual information (MI) metric in rigid-body registration of two digital images, in particular, local fluctuations of the MI value due to interpolation. In contrast to existing work in this area, this paper starts with two hypothetical continuous images, based on which both sampling and interpolation effects are analyzed. This analysis indicates that an "ideal" interpolator may not be able to completely suppress the undesirable local minima of the MI metric if the sampling effect is not negligible. Several preprocessing methods are discussed for reducing the interpolation effects.

Many imaging experiments involve acquiring a time series of images. To improve imaging speed, several "data-sharing" methods have been proposed, which collect one (or a few) high-resolution reference(s) and a sequence of reduced data sets. In image reconstruction, two methods, known as "Keyhole" and reduced-encoding imaging by generalized-series reconstruction (RIGR), have been used. Keyhole fills in the unmeasured high-frequency data simply with those from the reference data set(s), whereas RIGR recovers the unmeasured data using a generalized series (GS) model, of which the basis functions are constructed based on the reference image(s). This correspondence presents a fast algorithm (and two extensions) for GS-based image reconstruction. The proposed algorithms have the same computational complexity as the Keyhole algorithm, but are more capable of capturing high-resolution dynamic signal changes.

Guest Editorial Technological advances in biomedical imaging are providing unprecedented opportunities for improving our understanding of biological processes and revealing the anatomical and functional organization of biological systems, from macro- to nano-scales. Research in this area is progressing at an extraordinary rate and is becoming more and more interdisciplinary. The IEEE International Symposium on Biomedical Imaging (ISBI) was initiated to bring together researchers from the medical and biological imaging communities and to provide an effective forum for multidisciplinary interactions. The first meeting was held at the Ritz-Carlton Hotel, Washington, DC, July 7-10, 2002. It was organized jointly by the IEEE Signal Processing Society (SPS) and the IEEE Engineering in Medicine and Biology Society (EMBS) and co-sponsored by National Institutes of Health's (NIH) National Institute of the Biomedical Imaging and Bioengineering (NIBIB), representing the first IEEE-NIH collaborative effort on a major imaging conference. The symposium was focused on the engineering aspects of biomedical imaging while promoting an integrative approach through all scales of observation. It successfully brought together a large group of biomedical imaging researchers and practitioners (535 participants) with different backgrounds to share their knowledge and to address the latest challenges in data acquisition, image reconstruction, image processing, analysis, and visualization. A highlight of the symposium was the inspired opening address by Dr. Elias Zerhouni, the newly appointed director of the NIH. He stressed the importance of imaging for the biomedical sciences and expressed a strong interest in the conference. Dr. Zerhouni is very much in favor of collaborations between engineers and biomedical scientists, having experienced them first hand as a radiologist. He holds several patents related to imaging and is widely known for having introduced magnetic resonance tagging as a diagnostic tool for assessing heart function. Another noted NIH speaker was Dr. Roderic Pettigrew, who will become NIBIB's first permanent director in September 2002. He opened the first session by expressing his strong support for ISBI; he also explained the mission of NIBIB in which imaging will play a major role. The scientific program of the conference consisted of 3 plenary talks, 10 special sessions, 19 oral sessions, and 6 poster sessions over three days. The plenary talks, one on each day, covered topics from molecular imaging (by Michael Phelps, University of California, Los Angeles (UCLA), School of Medicine), functional imaging (by Alan Koretsky, NIH), to medical image analysis (by Michael Brady, University of Oxford, U.K.). The first two talks reviewed and discussed the latest developments in imaging modalities and applications, while the latter illustrated the power and potential of image processing and analysis algorithms. The special sessions, consisting of invited papers and sponsored by NIBIB, were a unique feature of the conference. They covered a wide range of topics: "Model-Based Image Segmentation and Analysis" (organized by Christos Davatzikos), "Microarray Image Processing and Analysis" (organized by Bin Yu and Dan Bartell), "Optical Coherence Tomography" (organized by Stephen Boppart and René Salathé), "Micro Imaging" (organized by Erik Ritman and Françoise Peyrin), "Brain Connectivity and Functional Assessment" (organized by Peter Basser and John George), "Nonrigid Registration" (organized by Benoit Dawant), "Fast Acquisition and Sampling in MRI" (organized by Yoram Bresler), "Electron Microscopy" (organized by Benes Trus), "In Vivo Cellular and Molecular Imaging" (organized by John Hoffman and Gary Kelloff), and "Federal Funding Opportunities for Biomedical Imaging Research" (organized by Richard Swaja, NIH/NIBIB). In addition to the invited program, a total of 355 papers were submitted to the symposium, of which 73 were accepted for oral presentation and 142 for poster presentation. The acceptance rate for contributed papers was about 60%. The papers of the symposium are published in IEEE conference proceedings; they will be included in the IEEEXplore database which is searchable through the WEB. There are many indications that ISBI fulfills a strong need of the biomedical imaging community. Many participants expressed enthusiastic support for the conference. It was felt that the time was ripe for IEEE to have its flagship conference on biomedical imaging. In charge of planning the next meeting are Christian Roux (representing EMBS) and Richard Leahy (representing SPS). It is our belief and hope that the next ISBI, to be held in Washington, DC, once again, will be even more successful in serving our community. Efforts will be made to increase the participation and the level of cross fertilization between the medical and biological imaging communities. ISBI may also play an important role in graduate education and training, as it is often impossible for a single institution to develop a comprehensive biomedical imaging curriculum covering all the modalities and applications. It is, therefore, desirable that future ISBI's include a one-day pre-conference event with tutorials and short courses covering not only imaging fundamentals, but also emerging technologies and applications. In closing, we may say that ISBI was a success, partly due to timing, location, and the fact that the conference responded to a need of the engineering community. However, nothing could have happened without the hard work, creativity, and involvement of the volunteers. We would like to take this opportunity to thank the entire organizing committee, the program committee, the special session chairs, and the external reviewers for their large commitment of time and effort. All administrative matters were handled by the staff of both SPS and EMBS under the leadership of their executive directors, Mercy Kowalczyk and Laura Wolf. The technical program chairs were Jeffrey Fessler (for SPS) and Michael Vannier (for EMBS), with additional support provided by the special sessions chairs Erik Meijering and Jean-Louis Coatrieux and the NIBIB representative on the committee, Richard Swaja, who played a crucial role in involving the NIH. We do not know how to thank them enough for having helped us to turn what initially looked like a nice idea into a reality.

This paper presents a new method for image segmentation by deforming the object shape in a template. The deformation process is controlled using a thin-plate spline kernel based regularization method. The proposed method is especially useful for 2D-based segmentation of 3D medical images by treating segmented slices as templates for their neighboring unsegmented slices. We have applied the proposed method to extract the scalp contours in brain cryosection images with very encouraging results.

This paper presents a novel landmark-based shape deformation method. This method effectively solves two problems inherent in landmark-based shape deformation: (a) identification of landmark points from a given input image, and (b) regularized deformation the shape of an an object defined in a template. The second problem is solved using a new constrained support vector machine (SVM) regression technique, in which a thin-plate kernel is utilized to provide non-rigid shape deformations. This method offers several advantages over existing landmark-based methods. First, it has a unique capability to detect and use multiple candidate landmark points in an input image to improve landmark detection. Second, it can handle the case of missing landmarks, which often arises in dealing with occluded images. We have applied the proposed method to extract the scalp contours from brain cryosection images with very encouraging results.

An upper bound is derived for the mutual information between a fixed image and a deformable template containing a fixed number of gray-levels. The bound can be calculated by maximizing the entropy of the template under the constraint that the conditional entropy of the template, given the fixed image, be zero. This bound provides useful insight into the properties of mutual information as a similarity metric for deformable image registration. Specifically, it indicates that maximizing mutual information may not necessarily produce an optimal solution when the deformable transform is too flexible.

This paper describes a technique for using prior knowledge of shape variations to help guide a volumetric deformation process. The volumetric transform maintains the topology of a template while matching the template to an image under study. A statistical model is used to describe inter- and intra-shape correlations in the template. The parameters for the shape model are learned by performing eigenshape analysis on a training set consisting of deformations of a single template to several typical segmentations. The shape model is used to guide the deformation by the inclusion of a term to the deformation cost functional that promotes the most likely deformations according to the shape priors. Some advantages of the proposed method are that it inherently conserves the topology between multiple shapes, and that prelabeling of corresponding points and point ordering of the training set is not needed. Results are presented for segmentation of magnetic resonance and cryosection images with varying contrasts. A qualitative analysis shows that the inclusion of shape priors can significantly improve the final deformation result.

In this paper, the audio and visual features of speech are integrated using a novel fused-HMM. We assume that the two sets of features may have different data rates and duration. Hidden Markov models (HMMs) are first used to model them separately, and then a general Bayesian fusion method, which is optimal in the maximum entropy sense, is employed to fuse them together. Particularly, an efficient learning algorithm is introduced. Instead of maximizing the joint likelihood of the fuse-HMM, the learning algorithm maximizes the two HMMs separately, and then fuses the HMMs together. In addition, an inference algorithm is proposed. We have tested the proposed method by person verification experiments. Results show that the proposed method significantly reduces the recognition error rates as compared to the unimodal HMMs and the loosely-coupled fusion model.

This paper presents a novel approach for multisensory information fusion in the Bayesian inference framework. Specifically, under the maximum entropy principle, a formula is derived for estimating the joint probabilities of multisensory signals. The formula uses appropriate mapping functions to reflect the dependencies among multisensory signals. Selection of the mappings is guided by the maximum mutual information criterion. In addition, an algorithm is proposed for linear mappings of Gaussian random variables. Experiments on simulated Gaussian data and video/audio signals have been carried out. Preliminary results demonstrate that the proposed method can significantly improve the recognition accuracy for this type of tasks.

This paper presents a novel method for the detection of neuronal activity-dependent signal changes in functional magnetic resonance imaging (fMRI) image sequences. In this method, wavelet analysis and statistical testing are applied jointly, enabling fMRI image sequences to be effectively analyzed in a unique spatio temporal framework. Experimental results show that this method performs significantly better than several existing methods for fMRI data with low SNR.

This paper addresses the classical image reconstruction problem from limited Fourier data. Here, we assume that a high-resolution reference which provides an initial estimate of the desired image is available. A new algorithm is described which represents the desired image using a family of basis functions derived from the reference image. The selection of the most efficient basis function set from this family is guided by the principle of maximum cross-entropy. Simulation and experimental results have shown that the algorithm can achieve high resolution with a small number of data points and can also account for relative rotation and translation between the reference and the measured data.

This paper discusses a novel technique for information fusion. Specifically, a formula is derived for estimation of the joint probabilities in the maximum entropy sense. In addition, neural networks are used to estimate conditional probabilities required in the Bayesian inference method. Preliminary experimental results demonstrate that the proposed method can significantly improve the accuracy of the bimodal recognition system using audio/video signals.

This paper presents a robust algorithm for automated registration of images related by rigid-body transformations. This algorithm uses a new region-based similarity metric, which enables accurate registration of images of large contrast differences. Region segmentation required by the metric is accomplished using a multiscale segmentation algorithm, and minimization of this metric is done using the Powell direction set method. Experimental results are presented to demonstrate that the algorithm is effective for aligning images from single or multiple imaging modalities without the use of any fiducial markers.

Neuronet approaches offer a unique and powerful tool for nonlinear information fusion. Unlike traditional techniques, neuronets do not require explicit environmental models or descriptions of sensor characteristics. This paper describes a technique for sensor fusion which makes use of a new neural model to combine data autonomously extracted from different sources. Application of the technique to bimodal recognition of combined speech/image signals is discussed.

This article formally defines partial Radon transforms for functions of more than two dimensions. It shows that a generalized projection-slice theorem exists which connects planar and hyperplanar projections of a function to its Fourier transform. In addition, a general theoretical framework is provided for carrying out n-dimensional backprojection reconstruction in a multistage fashion through the use of the partial Radon transform.

Presents a model-based phase unwrapping method which represents the unwrapped phase function by a truncated Taylor series and a residual function. An efficient, noniterative computational algorithm is also proposed for calculating the model parameters from the phase derivatives. Sample experimental results are shown to demonstrate the effectiveness of the algorithm for extracting unwrapped phase images from two-dimensional (2-D) magnetic resonance imaging (MRI) data.

This paper presents a novel dynamic imaging method. This method models the object by a time-varying function, thus converting the dynamic imaging problem to a parameter identification problem. Experimental results demonstrate that this method can produce time-sequential images from a time-varying object with both high temporal and spatial resolution. The proposed method has been validated in magnetic resonance imaging by computer simulation, phantom study and animal study.

It is well known that the inverse radon transform exists in two forms for odd and even dimension functions, respectively. This note shows that these formulas become zero when the dimension requirement is violated, which results in a simple unified formula for the inverse radon transform of arbitrary dimensions.

Magnetic resonance imaging and spectroscopy techniques are now being used to detect changes in blood flow, volume and oxygenation level associated with brain function. The authors describe a prototype system, called a neuroscope, that provides a real-time acquisition, control and processing environment for functional brain studies. Preliminary experiments have shown that oxygenation sensitive changes in the rat can be captured in real-time. When fully developed, this system should prove very useful for mapping the spatial and temporal patterns of functional brain activity of humans.>

Many magnetic resonance imaging applications require the acquisition of a time series of images. In conventional Fourier transform based imaging methods, each of these images is acquired independently so that the temporal resolution possible is limited by the number of spatial encodings (or data points in the Fourier space) collected, or one has to sacrifice spatial resolution for temporal resolution. Here, a generalized series based imaging technique is proposed to address this problem. This technique makes use of the fact that, in most time-sequential imaging problems, the high-resolution image morphology does not change from one image to another, and it improves imaging efficiency (and temporal resolution) over the conventional Fourier imaging methods by eliminating the repeated encodings of this stationary information. Additional advantages of the proposed imaging technique include a reduced number of radio frequency (RF) pulses for data collection, and thus lower RF power deposition. This method should prove useful for a variety of dynamic imaging applications, including dynamic studies of contrast agents and functional brain imaging.