Investigating the enhancement of template-free activation detection of event-related fMRI data using wavelet shrinkage and figures of merit.

OBJECTIVE: We propose a method for preprocessing event-related functional magnetic resonance imaging (fMRI) data that can lead to enhancement of template-free activation detection. The method is based on using a figure of merit to guide the wavelet shrinkage of a given fMRI data set. BACKGROUND: Several previous studies have demonstrated that in the root-mean-square error setting, wavelet shrinkage can improve the signal-to-noise ratio of fMRI time courses. However, preprocessing fMRI data in the root-mean-square error setting does not necessarily lead to enhancement of template-free activation detection. Motivated by this observation, in this paper, we move to the detection setting and investigate the possibility of using wavelet shrinkage to enhance template-free activation detection of fMRI data. METHODOLOGY: The main ingredients of our method are (i) forward wavelet transform of the voxel time courses, (ii) shrinking the resulting wavelet coefficients as directed by an appropriate figure of merit, (iii) inverse wavelet transform of the shrunk data, and (iv) submitting these preprocessed time courses to a given activation detection algorithm. Two figures of merit are developed in the paper, and two other figures of merit adapted from the literature are described. RESULTS: Receiver-operating characteristic analyses with simulated fMRI data showed quantitative evidence that data preprocessing as guided by the figures of merit developed in the paper can yield improved detectability of the template-free measures. We also demonstrate the application of our methodology on an experimental fMRI data set. CONCLUSIONS: The proposed method is useful for enhancing template-free activation detection in event-related fMRI data. It is of significant interest to extend the present framework to produce comprehensive, adaptive and fully automated preprocessing of fMRI data optimally suited for subsequent data analysis steps.

Improvement of spectral density-based activation detection of event-related fMRI data.

For event-related data obtained from an experimental paradigm with a periodic design, spectral density at the fundamental frequency of the paradigm has been used as a template-free activation detection measure. In this article, we build and expand upon this detection measure to create an improved, integrated measure. Such an integrated measure linearly combines information contained in the spectral densities at the fundamental frequency as well as the harmonics of the paradigm and in a spatial correlation function characterizing the degree of co-activation among neighboring voxels. Several figures of merit are described and used to find appropriate values for the coefficients in the linear combination. Using receiver-operating characteristic analysis on simulated functional magnetic resonance imaging (fMRI) data sets, we quantify and validate the improved performance of the integrated measure over the spectral density measure based on the fundamental frequency as well as over some other popular template-free data analysis methods. We then demonstrate the application of the new method on an experimental fMRI data set. Finally, several extensions to this work are suggested.

Scoring functions for de novo protein structure prediction revisited.

De novo protein structure prediction methods attempt to predict tertiary structures from sequences based on general principles that govern protein folding energetics and/or statistical tendencies of conformational features that native structures acquire, without the use of explicit templates. A general paradigm for de novo prediction involves sampling the conformational space, guided by scoring functions and other sequence-dependent biases, such that a large set of candidate ("decoy") structures are generated, and then selecting native-like conformations from those decoys using scoring functions as well as conformer clustering. High-resolution refinement is sometimes used as a final step to fine-tune native-like structures. There are two major classes of scoring functions. Physics-based functions are based on mathematical models describing aspects of the known physics of molecular interaction. Knowledge-based functions are formed with statistical models capturing aspects of the properties of native protein conformations. We discuss the implementation and use of some of the scoring functions from these two classes for de novo structure prediction in this chapter.

A knowledge-based scoring function based on residue triplets for protein structure prediction.

One of the general paradigms for ab initio protein structure prediction involves sampling the conformational space such that a large set of decoy (candidate) structures are generated and then selecting native-like conformations from those decoys using various scoring functions. In this study, based on a physical/geometric approach first suggested by Banavar and colleagues, we formulate a knowledge-based scoring function, which uses the radii of curvature formed among triplets of residues in a protein conformation. By analyzing its performance on various decoy sets, we determine a good set of parameters--the distance cutoff and the number of distance bins--to use for configuring such a function. Furthermore, we investigate the effect of using various approaches for compiling the prior distribution on the performance of the knowledge-based function. Possible extensions to the current form of the residue triplet scoring function are discussed.

PROTINFO: new algorithms for enhanced protein structure predictions.

We describe new algorithms and modules for protein structure prediction available as part of the PROTINFO web server. The modules, comparative and de novo modelling, have significantly improved back-end algorithms that were rigorously evaluated at the sixth meeting on the Critical Assessment of Protein Structure Prediction methods. We were one of four server groups invited to make an oral presentation (only the best performing groups are asked to do so). These two modules allow a user to submit a protein sequence and return atomic coordinates representing the tertiary structure of that protein. The PROTINFO server is available at http://protinfo.compbio.washington.edu.

Cluster significance testing using the bootstrap.

Many of the statistical methods currently employed to analyze fMRI data depend on a response template. However, the true form of the hemodynamic response, and thereby the response template, is often unknown. Consequently, cluster analysis provides a complementary, template-free method for exploratory analysis of multidimensional fMRI data sets. Clustering algorithms currently being applied to fMRI data separate the data into a predefined number of clusters (k). A poor choice of k will result in erroneously partitioning well-defined clusters. Although several clustering algorithms have been successfully applied to fMRI data, techniques for statistically testing cluster separation are still lacking. To address this problem we suggest a method based on Fisher's linear discriminant and the bootstrap. Also introduced in this paper is a measure based on the projection of multidimensional data from two clusters onto the vector, maximizing the ratio of the between- to the within-cluster sums of squares. The resulting one-dimensional distribution may be readily visualized and used as a heuristic for estimating cluster homogeneity. These methods are demonstrated for the self-organizing maps clustering algorithm when applied to event-related fMRI data.

Node merging in Kohonen's self-organizing mapping of fMRI data.

In this paper, Kohonen's self-organizing mapping (SOM) is used as a data-driven technique for analyzing functional magnetic resonance imaging (fMRI) data. Upon the completion of an SOM analysis, a cluster merging technique, based on examining the reproducibility of the fMRI data across epochs, is utilized to merge SOM nodes whose feature vectors are sufficiently similar to one another. The resulting 'super nodes' give time course templates of potential interest. These templates can be subsequently used in traditional template-based analysis methods, such as cross-correlation analysis, yielding statistical maps and activation patterns. This technique has been demonstrated on two fMRI datasets obtained from a visually-guided motor paradigm and a visual paradigm, respectively, showing satisfactory results.

Functional magnetic resonance imaging of the retina.

PURPOSE: This study explored the feasibility of mapping the retina's responses to visual stimuli noninvasively, by using functional magnetic resonance imaging (fMRI). METHODS: fMRI was performed on a 9.4-Tesla scanner to map activity-evoked signal changes of the retina-choroid complex associated with visual stimulation in anesthetized cats (n = 6). Three to 12 1-mm slices were acquired in a single shot using inversion-recovery, echo-planar imaging with a nominal in-plane resolution of 468 x 468 microm(2). Visual stimuli were presented to the full visual field and to the upper and lower visual fields. The stimuli were drifting or stationary gratings, which were compared with the dark condition. Activation maps were computed using cross-correlation analysis and overlaid on anatomic images. Multislice activation maps were reconstructed and flattened onto a two-dimensional surface. RESULTS: fMRI activation maps showed robust increased activity in the retina-choroid complex after visual stimulation. The average stimulus-evoked fMRI signal increase associated with drifting-grating stimulus was 1.7% +/- 0.5% (P < 10(-4), n = 6) compared with dark. Multislice functional images of the retina flattened onto a two-dimensional surface showed relatively uniform activation. No statistically significant activation was observed in and around the optic nerve head. Hemifield stimulation studies demonstrated that stimuli presented to the upper half of the visual field activated the lower part of the retina, and stimuli presented to the lower half of the visual field activated the upper part of the retina, as expected. Signal changes evoked by the stationary gratings compared with the dark basal condition were positive but were approximately half that evoked by the drifting gratings (1.0% +/- 0.1% versus 2.1% +/- 0.3%, P < 10(-4)). CONCLUSIONS: To the best of our knowledge, this is the first fMRI study of the retina, demonstrating its feasibility in imaging retinal function dynamically in a noninvasive manner and at relatively high spatial resolution.