Linear intensity normalization of FP-CIT SPECT brain images using the α-stable distribution.

In this work, a linear procedure to perform the intensity normalization of FP-CIT SPECT brain images is presented. This proposed methodology is based on the fact that the histogram of intensity values can be fitted accurately using a positive skewed α-stable distribution. Then, the predicted α-stable parameters and the location-scale property are used to linearly transform the intensity values in each voxel. This transformation is performed such that the new histograms in each image have a pre-specified α-stable distribution with desired location and dispersion values. The proposed methodology is compared with a similar approach assuming Gaussian distribution and the widely used specific-to-nonspecific ratio. In this work, we show that the linear normalization method using the α-stable distribution outperforms those existing methods.

Investigation of an efficient multi-modal convolutional neural network for multiple sclerosis lesion detection.

In this study, an automated 2D machine learning approach for fast and precise segmentation of MS lesions from multi-modal magnetic resonance images (mmMRI) is presented. The method is based on an U-Net like convolutional neural network (CNN) for automated 2D slice-based-segmentation of brain MRI volumes. The individual modalities are encoded in separate downsampling branches without weight sharing, to leverage the specific features. Skip connections input feature maps to multi-scale feature fusion (MSFF) blocks at every decoder stage of the network. Those are followed by multi-scale feature upsampling (MSFU) blocks which use the information about lesion shape and location. The CNN is evaluated on two publicly available datasets: The ISBI 2015 longitudinal MS lesion segmentation challenge dataset containing 19 subjects and the MICCAI 2016 MSSEG challenge dataset containing 15 subjects from various scanners. The proposed multi-input 2D architecture is among the top performing approaches in the ISBI challenge, to which open-access papers are available, is able to outperform state-of-the-art 3D approaches without additional post-processing, can be adapted to other scanners quickly, is robust against scanner variability and can be deployed for inference even on a standard laptop without a dedicated GPU.

Learning Intuitive Physics and One-Shot Imitation Using State-Action-Prediction Self-Organizing Maps.

Human learning and intelligence work differently from the supervised pattern recognition approach adopted in most deep learning architectures. Humans seem to learn rich representations by exploration and imitation, build causal models of the world, and use both to flexibly solve new tasks. We suggest a simple but effective unsupervised model which develops such characteristics. The agent learns to represent the dynamical physical properties of its environment by intrinsically motivated exploration and performs inference on this representation to reach goals. For this, a set of self-organizing maps which represent state-action pairs is combined with a causal model for sequence prediction. The proposed system is evaluated in the cartpole environment. After an initial phase of playful exploration, the agent can execute kinematic simulations of the environment's future and use those for action planning. We demonstrate its performance on a set of several related, but different one-shot imitation tasks, which the agent flexibly solves in an active inference style.

Dose voxel kernel prediction with neural networks for radiation dose estimation.

BACKGROUND: Currently there is an ever increasing interest in Lu-177 targeted radionuclide therapies, which target neuro-endocrine and prostate tumours. For a patient-specific treatment, an individual dosimetry based on SPECT/CT imaging is necessary. The aim of this study is to introduce a dosimetry method, where dose voxel kernels (DVK) are predicted by a neural network. METHODS: Kidneys are considered one of the most critical organs in any radionuclide therapy. Hence we chose kidneys of 26 patients, who underwent Lu-177-DOTATOC or PSMA therapy, as target organs for our dosimetric method. First of all, density kernels with a size of 9×9×9 voxels were considered, and the corresponding DVKs were calculated by Monte Carlo simulations. These kernels were used to train a neural network (NN), which received a density kernel as input and predicted a DVK at the output. To predict the dose distribution of an entire kidney, the organ had to be partitioned into a large number of density kernels. Afterwards the DVKs were predicted by a trained NN, and employed to reconstruct the whole-organ dose distribution by convolution with the activity distribution. This method was compared to the standard method where the activity distribution is convolved with a DVK based on a homogeneous soft tissue kernel. RESULTS: The number of training kernels amounted to 52,274 density kernels with corresponding MC-derived DVKs. The results serve as proof of principle of the newly proposed method and showed that the NN approach yielded superior results compared to the standard method with no additional computational effort. CONCLUSION: The NN approach is an accurate and highly competitive dosimetric method to precisely estimate absorbed radiation dose in critical organs like kidneys in clinical routine. To further improve the results, a larger number of DVKs needs to be computed by Monte Carlo simulations. An extension of the method to other organs is easily conceivable.

Brain Connectivity Studies on Structure-Function Relationships: A Short Survey with an Emphasis on Machine Learning.

This short survey reviews the recent literature on the relationship between the brain structure and its functional dynamics. Imaging techniques such as diffusion tensor imaging (DTI) make it possible to reconstruct axonal fiber tracks and describe the structural connectivity (SC) between brain regions. By measuring fluctuations in neuronal activity, functional magnetic resonance imaging (fMRI) provides insights into the dynamics within this structural network. One key for a better understanding of brain mechanisms is to investigate how these fast dynamics emerge on a relatively stable structural backbone. So far, computational simulations and methods from graph theory have been mainly used for modeling this relationship. Machine learning techniques have already been established in neuroimaging for identifying functionally independent brain networks and classifying pathological brain states. This survey focuses on methods from machine learning, which contribute to our understanding of functional interactions between brain regions and their relation to the underlying anatomical substrate.

Intronic microRNAs support their host genes by mediating synergistic and antagonistic regulatory effects.

BACKGROUND: MicroRNA-mediated control of gene expression via translational inhibition has substantial impact on cellular regulatory mechanisms. About 37% of mammalian microRNAs appear to be located within introns of protein coding genes, linking their expression to the promoter-driven regulation of the host gene. In our study we investigate this linkage towards a relationship beyond transcriptional co-regulation. RESULTS: Using measures based on both annotation and experimental data, we show that intronic microRNAs tend to support their host genes by regulation of target gene expression with significantly correlated expression patterns. We used expression data of three differentiating cell types and compared gene expression profiles of host and target genes. Many microRNA target genes show expression patterns significantly correlated with the expressions of the microRNA host genes. By calculating functional similarities between host and predicted microRNA target genes based on GO annotations, we confirm that many microRNAs link host and target gene activity in an either synergistic or antagonistic manner. CONCLUSIONS: These two regulatory effects may result from fine tuning of target gene expression functionally related to the host or knock-down of remaining opponent target gene expression. This finding allows to extend the common practice of mapping large scale gene expression data to protein associated genes with functionality of co-expressed intronic microRNAs.

Automated solvent artifact removal and base plane correction of multidimensional NMR protein spectra by AUREMOL-SSA.

Strong solvent signals lead to a disappearance of weak protein signals close to the solvent resonance frequency and to base plane variations all over the spectrum. AUREMOL-SSA provides an automated approach for solvent artifact removal from multidimensional NMR protein spectra. Its core algorithm is based on singular spectrum analysis (SSA) in the time domain and is combined with an automated base plane correction in the frequency domain. The performance of the method has been tested on synthetic and experimental spectra including two-dimensional NOESY and TOCSY spectra and a three-dimensional (1)H,(13)C-HCCH-TOCSY spectrum. It can also be applied to frequency domain spectra since an optional inverse Fourier transformation is included in the algorithm.

A Constrained ICA-EMD Model for Group Level fMRI Analysis.

Independent component analysis (ICA), being a data-driven method, has been shown to be a powerful tool for functional magnetic resonance imaging (fMRI) data analysis. One drawback of this multivariate approach is that it is not, in general, compatible with the analysis of group data. Various techniques have been proposed to overcome this limitation of ICA. In this paper, a novel ICA-based workflow for extracting resting-state networks from fMRI group studies is proposed. An empirical mode decomposition (EMD) is used, in a data-driven manner, to generate reference signals that can be incorporated into a constrained version of ICA (cICA), thereby eliminating the inherent ambiguities of ICA. The results of the proposed workflow are then compared to those obtained by a widely used group ICA approach for fMRI analysis. In this study, we demonstrate that intrinsic modes, extracted by EMD, are suitable to serve as references for cICA. This approach yields typical resting-state patterns that are consistent over subjects. By introducing these reference signals into the ICA, our processing pipeline yields comparable activity patterns across subjects in a mathematically transparent manner. Our approach provides a user-friendly tool to adjust the trade-off between a high similarity across subjects and preserving individual subject features of the independent components.

Particle filter de-noising of voxel-specific time-activity-curves in personalized 177Lu therapy.

BACKGROUND: Currently, there is a high interest in 177Lu targeted radionuclide therapies, which could be attributed to favorable results obtained from 177Lu compounds targeting neuro-endocrine and prostate tumors. SPECT based dosimetry could be used for deriving dose values for individual voxels, as is the standard in external-beam radiation-therapy (EBRT). For this a time-activity-curve (TAC) at voxel resolution and also a voxel-wise modeling of radiation energy deposition are necessary. But a voxel-wise determination of TACs is problematic, since several confounding factors exist, such as e.g. poor count-statistics or registration inaccuracies, which add noise to the observed activity states. A particle filter (PF) is a class of methods which applies regularization based on a model of the temporal evolution of activity states. The aim of this study is to introduce the application of PFs for de-noising of per-voxel time-activity curves. METHODS: We applied a PF for de-noising the TACs of 26 patients, who underwent 177Lu-DOTATOC or -PSMA therapy. The TACs were obtained from fully-quantitative, serial SPECT(/CT) data, acquired at 4h, 24h, 48h, 72h p.i. The model used in the PF was a mono-exponential decay and its free parameters were determined based on objective criteria. The time-integrated activities (TIA) resulting from the PF (PFF) were compared to the results of a mono-exponential fit (SF) of individual voxels in several volumes of interest (kidneys, spleen, tumors). Additionally, an organ-averaged TIA was derived from whole-organ VOIs and subsequent curve-fitting. This whole-organ TIA was also compared to the whole-organ TIAs obtained from summation of the voxel-wise TIAs from PFF and SF. RESULTS: The number of particles was set to 1000. Optimal values for noise of observations and noise of the model were 0.25 and 0.5, respectively. The deviation of whole-organ TIAs from conventional organ-based dosimetry and the summation of the voxel-wise TIAs was substantial for SF (kidneys -22.3%, spleen -49.6%, tumor -60.0%), as well as for PFF (kidneys -37.1%, spleen -57.9%, tumor -70.9%). The distribution of voxel-wise half-lives resulting from the PFF method was considerably closer to the organ-averaged value, and the number of implausibly long half-lives (>physical HL) was reduced. CONCLUSION: The PFF leads to voxel-wise half-lives, which are more plausible than those resulting from SF. However, one has to admit that voxel-wise fitting generally leads to considerable deviations from the organ-averaged TIA as obtained by conventional whole-organ evaluation. Unfortunately, we did not have ground-truth TIA of our patient data and proper ground-truth could even be impossible to obtain. Nevertheless, there are strong indicators that particle filtering can be used for reducing voxel-wise TAC noise.

Analyzing M-CSF dependent monocyte/macrophage differentiation: expression modes and meta-modes derived from an independent component analysis.

BACKGROUND: The analysis of high-throughput gene expression data sets derived from microarray experiments still is a field of extensive investigation. Although new approaches and algorithms are published continuously, mostly conventional methods like hierarchical clustering algorithms or variance analysis tools are used. Here we take a closer look at independent component analysis (ICA) which is already discussed widely as a new analysis approach. However, deep exploration of its applicability and relevance to concrete biological problems is still missing. In this study, we investigate the relevance of ICA in gaining new insights into well characterized regulatory mechanisms of M-CSF dependent macrophage differentiation. RESULTS: Statistically independent gene expression modes (GEM) were extracted from observed gene expression signatures (GES) through ICA of different microarray experiments. From each GEM we deduced a group of genes, henceforth called sub-mode. These sub-modes were further analyzed with different database query and literature mining tools and then combined to form so called meta-modes. With them we performed a knowledge-based pathway analysis and reconstructed a well known signal cascade. CONCLUSION: We show that ICA is an appropriate tool to uncover underlying biological mechanisms from microarray data. Most of the well known pathways of M-CSF dependent monocyte to macrophage differentiation can be identified by this unsupervised microarray data analysis. Moreover, recent research results like the involvement of proliferation associated cellular mechanisms during macrophage differentiation can be corroborated.

Frequency-Resolved Dynamic Functional Connectivity Reveals Scale-Stable Features of Connectivity-States.

Investigating temporal variability of functional connectivity is an emerging field in connectomics. Entering dynamic functional connectivity by applying sliding window techniques on resting-state fMRI (rs-fMRI) time courses emerged from this topic. We introduce frequency-resolved dynamic functional connectivity (frdFC) by means of multivariate empirical mode decomposition (MEMD) followed up by filter-bank investigations. In general, we find that MEMD is capable of generating time courses to perform frdFC and we discover that the structure of connectivity-states is robust over frequency scales and even becomes more evident with decreasing frequency. This scale-stability varies with the number of extracted clusters when applying k-means. We find a scale-stability drop-off from k = 4 to k = 5 extracted connectivity-states, which is corroborated by null-models, simulations, theoretical considerations, filter-banks, and scale-adjusted windows. Our filter-bank studies show that filter design is more delicate in the rs-fMRI than in the simulated case. Besides offering a baseline for further frdFC research, we suggest and demonstrate the use of scale-stability as a possible quality criterion for connectivity-state and model selection. We present first evidence showing that connectivity-states are both a multivariate, and a multiscale phenomenon. A data repository of our frequency-resolved time-series is provided.

Combined EMD-sLORETA Analysis of EEG Data Collected during a Contour Integration Task.

Lately, Ensemble Empirical Mode Decomposition (EEMD) techniques receive growing interest in biomedical data analysis. Event-Related Modes (ERMs) represent features extracted by an EEMD from electroencephalographic (EEG) recordings. We present a new approach for source localization of EEG data based on combining ERMs with inverse models. As the first step, 64 channel EEG recordings are pooled according to six brain areas and decomposed, by applying an EEMD, into their underlying ERMs. Then, based upon the problem at hand, the most closely related ERM, in terms of frequency and amplitude, is combined with inverse modeling techniques for source localization. More specifically, the standardized low resolution brain electromagnetic tomography (sLORETA) procedure is employed in this work. Accuracy and robustness of the results indicate that this approach deems highly promising in source localization techniques for EEG data.

Ensemble Empirical Mode Decomposition Analysis of EEG Data Collected during a Contour Integration Task.

We discuss a data-driven analysis of EEG data recorded during a combined EEG/fMRI study of visual processing during a contour integration task. The analysis is based on an ensemble empirical mode decomposition (EEMD) and discusses characteristic features of event related modes (ERMs) resulting from the decomposition. We identify clear differences in certain ERMs in response to contour vs noncontour Gabor stimuli mainly for response amplitudes peaking around 100 [ms] (called P100) and 200 [ms] (called N200) after stimulus onset, respectively. We observe early P100 and N200 responses at electrodes located in the occipital area of the brain, while late P100 and N200 responses appear at electrodes located in frontal brain areas. Signals at electrodes in central brain areas show bimodal early/late response signatures in certain ERMs. Head topographies clearly localize statistically significant response differences to both stimulus conditions. Our findings provide an independent proof of recent models which suggest that contour integration depends on distributed network activity within the brain.

Building a FP-CIT SPECT Brain Template Using a Posterization Approach.

Spatial affine registration of brain images to a common template is usually performed as a preprocessing step in intersubject and intrasubject comparison studies, computer-aided diagnosis, region of interest selection and brain segmentation in tomography. Nevertheless, it is not straightforward to build a template of [123I]FP-CIT SPECT brain images because they exhibit very low intensity values outside the striatum. In this work, we present a procedure to automatically build a [123I]FP-CIT SPECT template in the standard Montreal Neurological Institute (MNI) space. The proposed methodology consists of a head voxel selection using the Otsu's method, followed by a posterization of the source images to three different levels: background, head, and striatum. Analogously, we also design a posterized version of a brain image in the MNI space; subsequently, we perform a spatial affine registration of the posterized source images to this image. The intensity of the transformed images is normalized linearly, assuming that the histogram of the intensity values follows an alpha-stable distribution. Lastly, we build the [123I]FP-CIT SPECT template by means of the transformed and normalized images. The proposed methodology is a fully automatic procedure that has been shown to work accurately even when a high-resolution magnetic resonance image for each subject is not available.

Affine registration of [123I]FP-CIT SPECT brain images.

In this work, we perform a comparison between the spatial normalization of [123I]FP-CIT SPECT brain images when a FP-CIT SPECT and a MRI template are used. A 12-parameters affine registration model is calculated by the optimization of a sum of squares cost function. When the images are registered to a FP-CIT template, the intersubject variation is found to be lower than when the MRI template is used, specially in the striatum, which is the most relevant part of the brain in FP-CIT SPECT brain images.

Mathematical modeling of human brain physiological data.

Recently, a mathematical model of the basic physiological processes regulating the cerebral perfusion and oxygen supply was introduced [Jung et al., J. Math. Biol. 51, 491 (2005)]. Although this model correctly describes the interdependence of arterial blood pressure (ABP) and intracranial pressure (ICP), it fails badly when it comes to explaining certain abnormal correlations seen in about 80% of the recordings of ABP together with ICP and the partial oxygen pressure (TiPO(2)) of the neuronal tissue, taken at an intensive care unit during neuromonitoring of patients with a severe brain trauma. Such recordings occasionally show segments, where the mean arterial blood pressure is correlated with the partial oxygen pressure in tissue but anticorrelated with the intracranial pressure. The origin of such abnormal correlations has not been fully understood yet. Here, two extensions to the previous approach are proposed which can reproduce such abnormal correlations in simulations quantitatively. Furthermore, as the simulations are based on a mathematical model, additional insight into the physiological mechanisms from which such abnormal correlations originate can be gained.

Bilateral symmetry aspects in computer-aided Alzheimer's disease diagnosis by single-photon emission-computed tomography imaging.

OBJECTIVE: This paper explores the importance of the latent symmetry of the brain in computer-aided systems for diagnosing Alzheimer's disease (AD). Symmetry and asymmetry are studied from two points of view: (i) the development of an effective classifier within the scope of machine learning techniques, and (ii) the assessment of its relevance to the AD diagnosis in the early stages of the disease. METHODS: The proposed methodology is based on eigenimage decomposition of single-photon emission-computed tomography images, using an eigenspace extension to accommodate odd and even eigenvectors separately. This feature extraction technique allows for support-vector-machine classification and image analysis. RESULTS: Identification of AD patterns is improved when the latent symmetry of the brain is considered, with an estimated 92.78% accuracy (92.86% sensitivity, 92.68% specificity) using a linear kernel and a leave-one-out cross validation strategy. Also, asymmetries may be used to define a test for AD that is very specific (90.24% specificity) but not especially sensitive. CONCLUSIONS: Two main conclusions are derived from the analysis of the eigenimage spectrum. Firstly, the recognition of AD patterns is improved when considering only the symmetric part of the spectrum. Secondly, asymmetries in the hypo-metabolic patterns, when present, are more pronounced in subjects with AD.

Metal artifact reduction in cone beam computed tomography using forward projected reconstruction information.

In this work we present a new method to reduce artifacts, produced by high-density objects, especially metal implants, in X-ray cone beam computed tomography (CBCT). These artifacts influence clinical diagnostics and treatments using CT data, if metal objects are located in the field of view (FOV). Our novel method reduces metal artifacts by virtually replacing the metal objects with tissue objects of the same shape. First, the considered objects must be segmented in the original 2D projection data as well as in a reconstructed 3D volume. The attenuation coefficients of the segmented voxels are replaced with adequate attenuation coefficients of tissue (or water), then the required parts of the volume are projected onto the segmented 2D pixels, to replace the original information. This corrected 2D data can then be reconstructed with reduced artifacts, i. e. all metal objects virtually vanished. After the reconstruction, the segmented 3D metal objects were re-inserted into the corrected 3D volume. Our method was developed for mobile C-arm CBCTs; as it is necessary that they are of low weight, the C-arm results in unpredictable distortion. This misalignment between the original 2D data and the forward projection of the reconstructed 3D volume must be adjusted before the correction of the segmented 2D pixels. We applied this technique to clinical data and will now present the results.

Perceptual learning with perceptions.

UNLABELLED: In this work we present an approach to understand neuronal mechanisms underlying perceptual learning. Experimental results achieved with stimulus patterns of coherently moving dots are considered to build a simple neuronal model. The design of the model is made transparent and underlying behavioral assumptions made explicit. The key aspect of the suggested neuronal model is the learning algorithm used: We evaluated an implementation of Hebbian learning and are thus able to provide a straight-forward model capable to explain the neuronal dynamics underlying perceptual learning. Moreover, the simulation results suggest a very simple explanation for the aspect of "sub-threshold" learning (Watanabe et al. in Nature 413:844-884, 2001) as well as the relearning of motion discrimination after damage to primary visual cortex as recently reported (Huxlin et al. in J Neurosci 29:3981-3991, 2009) and at least indicate that perceptual learning might only occur when accompanied by conscious percepts. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (doi:10.1007/s11571-010-9134-9) contains supplementary material, which is available to authorized users.

Singular spectrum analysis for an automated solvent artifact removal and baseline correction of 1D NMR spectra.

NMR spectroscopy in biology and medicine is generally performed in aqueous solutions, thus in (1)H NMR spectroscopy, the dominant signal often stems from the partly suppressed solvent and can be many orders of magnitude larger than the resonances of interest. Strong solvent signals lead to a disappearance of weak resonances of interest close to the solvent artifact and to base plane variations all over the spectrum. The AUREMOL-SSA/ALS approach for automated solvent artifact removal and baseline correction has been originally developed for multi-dimensional NMR spectroscopy. Here, we describe the necessary adaptations for an automated application to one-dimensional NMR spectra. Its core algorithm is still based on singular spectrum analysis (SSA) applied on time domain signals (FIDs) and it is still combined with an automated baseline correction (ALS) in the frequency domain. However, both steps (SSA and ALS) have been modified in order to achieve optimal results when dealing with one-dimensional spectra. The performance of the method has been tested on one-dimensional synthetic and experimental spectra including the back-calculated spectrum of HPr protein and an experimental spectrum of a human urine sample. The latter has been recorded with the typically used NOESY-type 1D pulse sequence including water pre-saturation. Furthermore, the fully automated AUREMOL-SSA/ALS procedure includes the managing of oversampled, digitally filtered and zero-filled data and the correction of the frequency domain phase shift caused by the group delay time shift from the digital finite response filtering.