Dimensionality reduction of calcium-imaged neuronal population activity.

In the past few years, new technologies in the field of neuroscience have made it possible to simultaneously image activity in large populations of neurons at cellular resolution in behaving animals. In mid-2016, a huge repository of this so-called "calcium imaging" data was made publicly available. The availability of this large-scale data resource opens the door to a host of scientific questions for which new statistical methods must be developed. In this paper we consider the first step in the analysis of calcium imaging data-namely, identifying the neurons in a calcium imaging video. We propose a dictionary learning approach for this task. First, we perform image segmentation to develop a dictionary containing a huge number of candidate neurons. Next, we refine the dictionary using clustering. Finally, we apply the dictionary to select neurons and estimate their corresponding activity over time, using a sparse group lasso optimization problem. We assess performance on simulated calcium imaging data and apply our proposal to three calcium imaging data sets. Our proposed approach is implemented in the R package scalpel, which is available on CRAN.

Local appearance-based methods have been proposed recently for face recognition. We analyze the effects of different dimension reduction and normalization methods on local appearance-based face recognition in this paper. Each image is divided into equal sized blocks and six different dimension reduction methods are implemented for each block separately to create local visual feature vectors. On these local features, several normalization methods are applied in an attempt to eliminate the changes in lighting conditions and contrast differences among blocks of different face images. The experimental results show the improvements in recognition rates due to the effects of dimension reduction and normalization for three different classifiers. Usage of trainable dimension reduction methods instead of DCT and a new normalization method in our work (within-block normalization as referred in this paper) are two factors that makes difference from previous works in literature. The best performance is achieved using a block size of 16times16, performing dimension reduction using approximate pairwise accuracy criterion (aPAC) and applying within-block mean and variance normalization.

Neural data science: accelerating the experiment-analysis-theory cycle in large-scale neuroscience

This work presents a new interactive data visualization approach based on mixture of the outcomes of dimensionality reduction (DR) methods. Such a mixture is a weighted sum, whose weighting factors are defined by the user through a visual and intuitive interface. Additionally, the low-dimensional representation space produced by DR methods are graphically depicted using scatter plots powered via an interactive data-driven visualization. To do so, pairwise similarities are calculated and employed to define the graph to be drawn on the scatter plot. Our visualization approach enables the user to interactively combine DR methods while provided information about the structure of original data, making then the selection of a DR scheme more intuitive.

One of the dimension reduction (DR) methods for data-visualization, t-distributed stochastic neighbor embedding (t-SNE), has drawn increasing attention. t-SNE gives us better visualization than conventional DR methods, by relieving so-called crowding problem. The crowding problem is one of the curses of dimensionality, which is caused by discrepancy between high and low dimensional spaces. However, in t-SNE, it is assumed that the strength of the discrepancy is the same for all samples in all datasets regardless of ununiformity of distributions or the difference in dimensions, and this assumption sometimes ruins visualization. Here we propose a new DR method inhomogeneous t-SNE, in which the strength is estimated for each point and dataset. Experimental results show that such pointwise estimation is important for reasonable visualization and that the proposed method achieves better visualization than the original t-SNE.

Advances in functional magnetic resonance imaging (fMRI) have led to the ability to study the brain across many contexts. However, the large number of features generated by functional connectivity approaches may overfit the data. These problems can be overcome with either feature selection (FS) or dimensionality reduction (DR), which can be applied to less complex models. We utilize two open source datasets to compare the performance of DR/FS methods on cognitive task decoding using a suite of ML classifiers. While FS and DR methods have been used previously in decoding research, no systematic comparison of their performance has been undertaken. Here, we compare available methods using commonly utilized machine learning libraries to establish which methods provide the best predictive performance. We then conduct statistical tests to examine the relative contributions of DR and FS methods and classifiers on decoding accuracy. Neither DR or FS was found to be superior. However, differences were identified across datasets and tasks. In the majority of methods and datasets, a peak in predictive performance was found using a small percentage (005-.10%) of the total number of original features. Some methods perform better than the baseline method of prediction with all available features or selecting features randomly. Decoding performance with some datasets with some method exceeds that of deep learning approaches. These results suggest a "sweet spot" for the tradeoff between the retention of features and predictive accuracy.

This paper carries on research surrounding the influences produced by dimensionality reduction on machine learning classification effect. Firstly, paper constructs the analysis architecture of data dimension reduction classification, combines the two different unsupervised dimension reduction methods, locally linear embedding (LLE) and principal component analysis (PCA) with the five machine learning classification methods: Gradient Boosting Decision Tree (GBDT), Random Forest, Support Vector Machine (SVM), K-Nearest Neighbor (KNN) and Logistic Regression. And then uses the handwritten digital identification dataset to analyze the classification performance of these five classification methods on different dimension datasets by different dimensionality reduction methods. The analysis shows that using the appropriate dimensionality reduction method for dimensionality reduction classification can effectively improve the classification accuracy; the dimensionality reduction classification effect of non-linear dimensionality reduction method is generally better than the linear dimensionality reduction method; different machine learning classification algorithms have significant differences in the sensitivity of dimensions.

The main goal of monolingual models of sentence processing is to explain how the syntactic processor (or parser) assigns structure to an incoming string of words. The theoretical divide in the field has been about whether the architecture and mechanisms of the human sentence processor are modular—and computations are carried out serially—or whether it is interactive and computations are carried out in parallel. The debate about bilingual sentence processing has, instead, focused on whether bilingual speakers process their second language in a manner similar to monolingual speakers of the target language. Proposals rooted in generative approaches to language acquisition argue that adult second language (L2) learners lack access to the universal principles or the ability to reset parameters that guide language acquisition and language processing in their L2. Models grounded in neurocognitive approaches to memory hold that late bilinguals recruit different memory systems compared to native speakers of the target language. Other models have argued that differences in first and second language processing result largely from capacity differences, differences in susceptibility to interference, or lack of predictive ability. More recently, several studies have turned toward more experience-driven accounts, eschewing the earlier assumption that the first language is static and unchanging, and instead focusing on the interactive and interconnected nature of the bilingual linguistic system. These studies have revealed that not only does the first language (L1) affect L2 syntactic processing, but experience with the L2 can have ramifications for processing in the native language. A range of experimental techniques are employed to investigate how monolingual and bilingual speakers process language at the sentence level. Eye-tracking techniques allow measurement of responses, such as eye movements and pupil dilation, to study written and auditory language processing. Such measures permit insight into the cognitive processes that are engaged when individuals read written text or inspect visual scenes. Electrophysiological measures are particularly helpful for understanding the time course of neural activity associated with language and cognitive processes. Event-related potentials (ERPs) are obtained by recording and averaging across brain potentials associated with time-locked events (e.g., a word in a sentence). Electrophysiological measures are used to determine which stages of processing are affected by the experimental manipulation. Neuroimaging provides information about changes in brain structure and function. For example, the functional magnetic resonance imaging (fMRI) technique can be used to visualize which brain regions are engaged in processing a particular type of sentence. Although neuroimaging is a relatively-new methodology, it holds great promise for increasing our understanding of the dynamic processes in the brain related to language. The writing of this bibliography was supported in part by National Science Foundation (NSF) grant BCS-1535124 and OISE 1545900 to Paola E. Dussias.

Evidence is mixed whether fluid intelligence (Gf) is associated with increased or decreased alpha and beta band activity (7–30 Hz). Moreover, the Gf relationship with the delta and theta band activity (1–7 Hz) is unknown. We recorded electroencephalographic (EEG) data in 160 healthy adults solving Raven's Advanced Progressive Matrices with a randomized item order to control for item difficulty unaffected by sequential effects. The participants studied each matrix for 30 s before the response bank onset, so we could track the time course of neural activity during problem solving. We measured Gf using six tests. Gf positively correlated with the delta band power, while there was no correlation with the theta band power. For almost all of the participants, we identified the specific slow rhythm frequency, which varied in power as a function of item difficulty. We observed that the lower this frequency, the higher Gf, but only in men. Finally, the alpha and low-beta activity correlated negatively with Gf after we had filtered out the activity during idle intervals (the latter reflecting waiting for the response bank). Overall, the brain activity in the delta, alpha, and beta bands explained 22.6% of Gf variance.

BackgroundCongenital prosopagnosia is a severe face perception impairment which is not acquired by a brain lesion and is presumably present from birth. It manifests mostly by an inability to recognise familiar persons.Electrophysiological research has demonstrated the relevance to face processing of a negative deflection peaking around 170 ms, labelled accordingly as N170 in the electroencephalogram (EEG) and M170 in magnetoencephalography (MEG). The M170 was shown to be sensitive to the inversion of faces and to familiarity-two factors that are assumed to be crucial for congenital prosopagnosia. In order to locate the cognitive dysfunction and its neural correlates, we investigated the time course of neural activity in response to these manipulations.MethodologySeven individuals with congenital prosopagnosia and seven matched controls participated in the experiment. To explore brain activity with high accuracy in time, we recorded evoked magnetic fields (275 channel whole head MEG) while participants were looking at faces differing in familiarity (famous vs. unknown) and orientation (upright vs. inverted). The underlying neural sources were estimated by means of the least square minimum-norm-estimation (L2-MNE) approach.Principal FindingsThe behavioural data corroborate earlier findings on impaired configural processing in congenital prosopagnosia. For the M170, the overall results replicated earlier findings, with larger occipito-temporal brain responses to inverted than upright faces, and more right- than left-hemispheric activity. Compared to controls, participants with congenital prosopagnosia displayed a general decrease in brain activity, primarily over left occipitotemporal areas. This attenuation did not interact with familiarity or orientation.ConclusionsThe study substantiates the finding of an early involvement of the left hemisphere in symptoms of prosopagnosia. This might be related to an efficient and overused featural processing strategy which serves as a compensation of impaired configural processing.

During the COVID-19 pandemic, researchers have made efforts to detect COVID-19 through various methods. In the dataset used for this study, COVID-19 patients were identified using chest computed tomography (CT) images. High dimensionality is frequently an issue in machine learning image classification. Accordingly, this study implemented three dimensionality reduction methods in combination with various machine learning algorithms for improved classification. Principal component analysis (PCA), uniform manifold approximation and projection (UMAP), and diffusion maps were applied to the dataset to extract the most important features of the chest CT images. The extracted features were given as input either to logistic regression or the extreme gradient boosting (XGBoost) algorithm to perform classification. The strongest model identified from this study was diffusion maps in combination with logistic regression. This model, evaluated against existing models from similar studies in recent years, yielded strong performance for detecting COVID-19 cases using chest CT images. Our proposed model achieved 97.35% accuracy, 92.16% sensitivity, and 98.59% specificity on the held-out test set in differentiating between COVID-19-positive cases and healthy, non-COVID-19 cases. This study aimed to detect COVID-19 without the use of viral testing. Importantly, this method could assist clinicians in making an initial diagnosis, especially when viral testing is not available or timely enough for the patient’s case. This study also provides deeper insight into various dimensionality reduction methods and how compatible they are with biomedical imaging data. Models were trained using stratified cross-validation on the training set, with final performance evaluated on a held-out test set at the patient level to prevent data leakage. Additional imbalance-aware metrics were used to assess robustness given class distribution differences.

Previous studies have shown that epithelial Na+ channels (ENaCs) are activated by laminar shear stress (LSS). ENaCs with a high intrinsic open probability because of a mutation (betaS518K) or covalent modification of an introduced Cys residue (alphaS580C) in the pre-second transmembrane domain (pre-M2) were not activated by LSS, suggesting that the pre-M2 region participates in conformational rearrangements during channel activation. We examined the role of the pore region of the alpha-subunit in channel gating by studying the kinetics of activation by LSS of wild-type ENaC and channels with Cys mutations in the tract Ser576-Ser592. Whole cell Na+ currents were monitored in oocytes expressing wild-type or mutant ENaCs prior to and following application of LSS. Following a 2.2-s delay, a monoexponential increase in Na+ currents was observed with a time constant (tau) of 8.1 s in oocytes expressing wild-type ENaC. Cys substitutions within the alpha-subunit in the tract Ser580-Ser589 resulted in: (i) a reduction (Ser580-Trp585, Gly587) or increase (Ser589) in delay times preceding channel activation by LSS, (ii) an increase (Gln581, Leu584, Trp585, Phe586, Ser588) or decrease (Ser589) in the rate of channel activation, or (iii) a decrease in the magnitude of the response (Ser583, Gly587, Leu584). Cys substitutions at a putative amiloride-binding site (alphaSer583 or betaGly525) or within the selectivity filter (alphaGly587) resulted in a reduction in the LSS response, and exhibited a multiexponential time course of activation. The corresponding gamma-subunit mutant (alphabetagammaG542C) had a minimal response to LSS and exhibited a high intrinsic open probability. These data suggest that residues in the pore region participate in the sensing and/or transduction of the mechanical stimulus that results in channel activation and are consistent with the hypothesis that the ENaC pore region has a key role in modulating channel gating.

Dimension reduction has been used to visualise the distribution of multidimensional microbiome data, but the composite variables calculated by the dimension reduction methods have not been widely used to investigate the relationship of the human gut microbiome with lifestyle and disease. In the present study, we applied several dimension reduction methods, including principal component analysis, principal coordinate analysis (PCoA), non-metric multidimensional scaling (NMDS), and non-negative matrix factorization, to a microbiome dataset from 186 subjects with symptoms of allergic rhinitis (AR) and 106 controls. All the dimension reduction methods supported that the distribution of microbial data points appeared to be continuous rather than discrete. Comparison of the composite variables calculated from the different dimension reduction methods showed that the characteristics of the composite variables differed depending on the distance matrices and the dimension reduction methods. The first composite variables calculated from PCoA and NMDS with the UniFrac distance were strongly associated with AR (FDR adjusted P = 2.4 × 10–4 for PCoA and P = 2.8 × 10–4 for NMDS), and also with the relative abundance of Bifidobacterium and Prevotella. The abundance of Bifidobacterium was also linked to intake of several nutrients, including carbohydrate, saturated fat, and alcohol via composite variables. Notably, the association between the composite variables and AR was much stronger than the association between the relative abundance of individual genera and AR. Our results highlight the usefulness of the dimension reduction methods for investigating the association of microbial composition with lifestyle and disease in clinical research.

Calcium imaging allows recording from hundreds of neurons in vivo with the ability to resolve single cell activity. Evaluating and analyzing neuronal responses, while also considering all dimensions of the data set to make specific conclusions, is extremely difficult. Often, descriptive statistics are used to analyze these forms of data. These analyses, however, remove variance by averaging the responses of single neurons across recording sessions, or across combinations of neurons, to create single quantitative metrics, losing the temporal dynamics of neuronal activity, and their responses relative to each other. Dimensionally Reduction (DR) methods serve as a good foundation for these analyses because they reduce the dimensions of the data into components, while still maintaining the variance. Nonnegative Matrix Factorization (NMF) is an especially promising DR analysis method for analyzing activity recorded in calcium imaging because of its mathematical constraints, which include positivity and linearity. We adapt NMF for our analyses and compare its performance to alternative dimensionality reduction methods on both artificial and in vivo data. We find that NMF is well-suited for analyzing calcium imaging recordings, accurately capturing the underlying dynamics of the data, and outperforming alternative methods in common use.

Understanding the neural basis of mosquito behavior is critical for designing effective vector control strategies and can potentially shed new light on basic nervous system function. Because mosquitoes are a non-model species, however, functional studies of mosquito nervous systems have long been restricted to electrophysiological recording from peripheral sensory organs such as the antenna. This is now changing with the advent of CRISPR-Cas9 gene editing and the development of other powerful new genetic tools. Transgenic mosquitoes that carry genetically encoded calcium sensors, for example, open the door to optical recording of neural activity with two-photon calcium imaging. Compared with electrophysiology, calcium imaging permits continuous monitoring of neural activity from large populations of neurons, even deep in the brain. When combined with selective neural drivers, it also allows targeted recording from specific neuronal types. Here, we describe a calcium imaging protocol we use in our laboratory to study neural activity in the brain of Aedes aegypti mosquitoes.