Classification using sparse representations: a biologically plausible approach

Representing signals as linear combinations of basis vectors sparsely selected from an overcomplete dictionary has proven to be advantageous for many applications in pattern recognition, machine learning, signal processing, and computer vision. While this approach was originally inspired by insights into cortical information processing, biologically plausible approaches have been limited to exploring the functionality of early sensory processing in the brain, while more practical applications have employed non-biologically plausible sparse coding algorithms. Here, a biologically plausible algorithm is proposed that can be applied to practical problems. This algorithm is evaluated using standard benchmark tasks in the domain of pattern classification, and its performance is compared to a wide range of alternative algorithms that are widely used in signal and image processing. The results show that for the classification tasks performed here, the proposed method is competitive with the best of the alternative algorithms that have been evaluated. This demonstrates that classification using sparse representations can be performed in a neurally plausible manner, and hence, that this mechanism of classification might be exploited by the brain.     

Interactions between landcover pattern and geospatial processing methods: Effects on landscape metrics and classification accuracy

Remote sensing data is routinely used in ecology to investigate the relationship between landscape pattern as characterised by land use and land cover maps, and ecological processes. Multiple factors related to the representation of geographic phenomenon have been shown to affect characterisation of landscape pattern resulting in spatial uncertainty. This study investigated the effect of the interaction between landscape spatial pattern and geospatial processing methods statistically; unlike most papers which consider the effect of each factor in isolation only. This is important since data used to calculate landscape metrics typically undergo a series of data abstraction processing tasks and are rarely performed in isolation. The geospatial processing methods tested were the aggregation method and the choice of pixel size used to aggregate data. These were compared to two components of landscape pattern, spatial heterogeneity and the proportion of landcover class area. The interactions and their effect on the final landcover map were described using landscape metrics to measure landscape pattern and classification accuracy (response variables). All landscape metrics and classification accuracy were shown to be affected by both landscape pattern and by processing methods. Large variability in the response of those variables and interactions between the explanatory variables were observed. However, even though interactions occurred, this only affected the magnitude of the difference in landscape metric values. Thus, provided that the same processing methods are used, landscapes should retain their ranking when their landscape metrics are compared. For example, highly fragmented landscapes will always have larger values for the landscape metric “number of patches” than less fragmented landscapes. But the magnitude of difference between the landscapes may change and therefore absolute values of landscape metrics may need to be interpreted with caution. The explanatory variables which had the largest effects were spatial heterogeneity and pixel size. These explanatory variables tended to result in large main effects and large interactions. The high variability in the response variables and the interaction of the explanatory variables indicate it would be difficult to make generalisations about the impact of processing on landscape pattern as only two processing methods were tested and it is likely that untested processing methods will potentially result in even greater spatial uncertainty.     

Nonlinear transient computation as a potential "kernel trick" in cortical processing

Evidence has been found for the presence of chaotic dynamics at all levels of the mammalian brain. This has led to some searching questions about the potential role that nonlinear dynamics may have in neural information processing. We propose that chaos equips the brain with the equivalent of a kernel trick for solving hard nonlinear problems. The approach presented, which is described as nonlinear transient computation, uses the dynamics of a well known chaotic attractor. The paper provides experimental results to show that this approach can be used to solve some challenging pattern recognition tasks. The paper also offers evidence to suggest that the efficacy of nonlinear transient computation for nonlinear pattern classification is dependent only on the generic properties of chaotic attractors and is not sensitive to the particular dynamics of specific sub-regions of chaotic phase space. If, as this work suggests, nonlinear transient computation is independent of the particulars of any given chaotic attractor, then it could be offered as a possible explanation of how the chaotic dynamics that have been observed in brain structures contribute to neural information processing tasks.     

Tactile motion and pattern processing assessed with high-field FMRI

Processing of motion and pattern has been extensively studied in the visual domain, but much less in the somatosensory system. Here, we used ultra-high-field functional magnetic resonance imaging (fMRI) at 7 Tesla to investigate the neuronal correlates of tactile motion and pattern processing in humans under tightly controlled stimulation conditions. Different types of dynamic stimuli created the sensation of moving or stationary bar patterns during passive touch. Activity in somatosensory cortex was increased during both motion and pattern processing and modulated by motion directionality in primary and secondary somatosensory cortices (SI and SII) as well as by pattern orientation in the anterior intraparietal sulcus. Furthermore, tactile motion and pattern processing induced activity in the middle temporal cortex (hMT+/V5) and in the inferior parietal cortex (IPC), involving parts of the supramarginal und angular gyri. These responses covaried with subjects' individual perceptual performance, suggesting that hMT+/V5 and IPC contribute to conscious perception of specific tactile stimulus features. In addition, an analysis of effective connectivity using psychophysiological interactions (PPI) revealed increased functional coupling between SI and hMT+/V5 during motion processing, as well as between SI and IPC during pattern processing. This connectivity pattern provides evidence for the direct engagement of these specialized cortical areas in tactile processing during somesthesis.     

基于表面肌电信号的人体动作识别算法研究进展

表面肌电信号(Surface Electromyography,sEMG)是通过相应肌群表面的传感器记录下来的一维时间序列非平稳生物电信号,不但反映了神经肌肉系统活动,对于反映相应动作肢体活动信息同样重要。而模式识别是肌电应用领域的基础和关键。为了在应用基于表面肌电信号模式识别中选取合适算法,本文拟对基于表面肌电信号的人体动作识别算法进行回顾分析,主要包括模糊模式识别算法、线性判别分析算法、人工神经网络算法和支持向量机算法。模糊模式识别能自适应提取模糊规则,对初始化规则不敏感,适合处理s EMG这样具有严格不重复的生物电信号;线性判别分析对数据进行降维,计算简单,但不适合大数据;人工神经网络可以同时描述训练样本输入输出的线性关系和非线性映射关系,可以解决复杂的分类问题,学习能力强;支持向量机处理小样本、非线性的高维数据优势明显,计算速度快。比较各方法的优缺点,为今后处理此类问题模式识别算法选取提供了参考和依据。     

Recurring discharge patterns in multiple spike trains

Neural networks are parallel processing structures that provide the capability to perform various pattern recognition tasks. A network is typically trained over a set of exemplars by adjusting the weights of the interconnections using a back propagation algorithm. This gradient search converges to locally optimal solutions which may be far removed from the global optimum. In this paper, evolutionary programming is analyzed as a technique for training a general neural network. This approach can yield faster, more efficient yet robust training procedures that accommodate arbitrary interconnections and neurons possessing additional processing capabilities.     

In search for significant cognitive features in Klinefelter syndrome through cross-species comparison of a supernumerary X chromosome

The behavioral characterization of animals that carry genetic disorder abnormalities in a controlled genetic and environmental background may be used to identify human deficits that are significant to understand underlying neurobiological mechanisms. Here, we studied whether previously reported object recognition impairments in mice with a supernumerary X chromosome relate to specific cognitive deficits in Klinefelter syndrome (47,XXY). We aimed to optimize face validity by studying temporal object recognition in human cognitive assays. Thirty-four boys with Klinefelter syndrome (mean age 12.01) were compared with 90 age-matched normal controls, on a broad range of visual object memory tasks, including tests for pattern and temporal order discrimination. The results indicate that subjects with Klinefelter syndrome have difficulty in the processing of visual object and pattern information. Visual object patterns seem difficult to discriminate especially when temporal information needs to be processed and reproduced. On the basis of cross-species comparison, we propose that impaired temporal processing of object pattern information is an important deficit in Klinefelter syndrome. The current study shows how cross-species behavioral characterization may be used as a starting point to understand the neurobiology of syndromal phenotypic expression. The features of this study may serve as markers for interventions in Klinefelter syndrome. Similar cross-species evaluations of standard mouse behavioral paradigms in different genetic contexts may be powerful tools to optimize genotype-phenotype relationships.     

Nucleotide deletion and P addition in V(D)J recombination: a determinant role of the coding-end sequence.

During V(D)J recombination, the coding ends to be joined are extensively modified. Those modifications, termed coding-end processing, consist of removal and addition of various numbers of nucleotides. We previously showed in vivo that coding-end processing is specific for each coding end, suggesting that specific motifs in a coding-end sequence influence nucleotide deletion and P-region formation. In this study, we created a panel of recombination substrates containing actual immunoglobulin and T-cell receptor coding-end sequences and dissected the role of each motif by comparing its processing pattern with those of variants containing minimal nucleotide changes from the original sequence. Our results demonstrate the determinant role of specific sequence motifs on coding-end processing and also the importance of the context in which they are found. We show that minimal nucleotide changes in key positions of a coding-end sequence can result in dramatic changes in the processing pattern. We propose that each coding-end sequence dictates a unique hairpin structure, the result of a particular energy conformation between nucleotides organizing the loop and the stem, and that the interplay between this structure and specific sequence motifs influences the frequency and location of nicks which open the coding-end hairpin. These findings indicate that the sequences of the coding ends determine their own processing and have a profound impact on the development of the primary B- and T-cell repertoires.     

资源型城市土地利用变化的景观过程响应——以黑龙江省大庆市为例

土地利用与土地覆盖变化直接或间接地改变着景观生态系统的结构和功能,进而影响地表各种景观过程。地表景观过程的改变,必然对人类赖以生存的生态环境产生程度不同的影响。黑龙江省大庆市作为典型的资源型城市(以石油、草场和耕地资源发展起来),是人地关系作用最强烈的区域之一,也是土地利用发生急剧变化且具有特色的区域。揭示该区域土地利用的时空变化特征和变化规律,探讨土地利用变化的景观生态趋势及对生态环境的可能影响等,对于资源型城市生态环境改善,产业结构调整及实现资源型城市可持续发展尤为重要。为此,基于大庆市1978,1988,1992,1996,2001年5个时段的MSS和TM影像,获取20余年不同时期的土地利用信息,以GIS技术为数据集成分析平台,计算土地利用转移概率矩阵,景观格局指数和主要景观组分的转移过程贡献率,以期从景观格局的角度定量刻画区域景观过程对土地利用变化的响应。结果表明,区域内耕地不断增加,草地持续减少,建设用地和林地逐渐扩大,而湿地则一直大幅度减少;耕地和草地是区域的主要景观类型,但二者的景观动态及过程却不相同,耕地的景观过程表现为空间集中性加强,而草地的破碎化程度加强,景观组分间的转移过程频繁,景观不稳定;对维持当地生态系统稳定起重要作用的湿地景观,其破碎化过程最为明显且面积和平均斑块面积都在迅速减少,这种景观过程将对当地生态环境变化造成负面影响。草地退化,湿地萎缩和功能减退,沙化和盐碱化日趋严重等生态过程,是土地利用变化的景观生态过程响应,恢复与建设生态脆弱的资源型城市的景观生态系统,须对景观尺度上的土地利用方式进行合理的调控,而这一调控的有效途径在于政策、体制的调整和观念的变化。     

The selectivity of cathepsin D suggests an involvement of the enzyme in the generation of T-cell epitopes

The selectivity of cathepsin D, a mammalian intracellular aspartyl proteinase involved in the degradation of endocytosed proteins, was studied. For this purpose, several proteins of known primary structure were subjected to mild proteolysis by the enzyme, and the preferentially cleaved peptide bonds were identified. Comparison of the primary structures around these sites indicates that cathepsin D shows a strong preference for peptide bonds within a distinct sequence pattern of amino acids extending over 7 residues. In general, this pattern is most likely to occur within amphipathic alpha-helical structures. These findings and their possible implications are discussed together with additional evidence suggesting an important role for cathepsin D in the processing of protein antigens, an essential step for their recognition by T-cells. Accordingly, it is proposed that the proteolytic activity of cathepsin D is crucial in selecting processing sites and hence the location and structural context of T-cell epitopes for the majority of protein antigens.     

Sex-Differences of Face Coding: Evidence from Larger Right Hemispheric M170 in Men and Dipole Source Modelling

The processing of faces relies on a specialized neural system comprising bilateral cortical structures with a dominance of the right hemisphere. However, due to inconsistencies of earlier findings as well as more recent results such functional lateralization has become a topic of discussion. In particular, studies employing behavioural tasks and electrophysiological methods indicate a dominance of the right hemisphere during face perception only in men whereas women exhibit symmetric and bilateral face processing. The aim of this study was to further investigate such sex differences in hemispheric processing of personally familiar and opposite-sex faces using whole-head magnetoencephalography (MEG). We found a right-lateralized M170-component in occipito-temporal sensor clusters in men as opposed to a bilateral response in women. Furthermore, the same pattern was obtained in performing dipole localization and determining dipole strength in the M170-timewindow. These results suggest asymmetric involvement of face-responsive neural structures in men and allow to ascribe this asymmetry to the fusiform gyrus. This specifies findings from previous investigations employing event-related potentials (ERP) and LORETA reconstruction methods yielding rather extended bilateral activations showing left asymmetry in women and right lateralization in men. We discuss our finding of an asymmetric fusiform activation pattern in men in terms of holistic face processing during face evaluation and sex differences with regard to visual strategies in general and interest for opposite faces in special. Taken together the pattern of hemispheric specialization observed here yields new insights into sex differences in face perception and entails further questions about interactions between biological sex, psychological gender and influences that might be stimulus-driven or task dependent.     

N-glycosylation modification of proteins is an early marker of the enterocytic differentiation process of HT-29 cells

The human colon cancer cell line HT-29 remains totally undifferentiated when glucose is present in the culture medium (HT-29 Glc+), while the same cells may undergo typical enterocytic differentiation after reaching confluence when grown in glucose-deprived medium (HT-29 Glc-). Recently, we demonstrated a deficiency in the overall N-glycan processing in confluent undifferentiated cells, whereas differentiated cells follow a classical pattern of N-glycosylation. The main changes in N-glycosylation observed in confluent undifferentiated cells may be summarised as follows: 1) the conversion of high mannose into complex glycopeptides is greatly decreased; 2) this decreased conversion could be a consequence of an accumulation of Man9-8-GlcNAc2-Asn high mannose species. Whether these changes in N-glycan processing appear progressively during cell culture or are already present from the beginning of the culture was investigated in this study by comparing the actual status of N-glycan processing in exponentially growing HT-29 Glc- and HT-29 Glc+ cells. Under these conditions, HT-29 Glc- cells do not exhibit any characteristics of differentiation. The conversion of high mannose into complex glycoproteins is severely reduced in HT-29 Glc+ cells, regardless of the growth phase studied. In contrast, HT-29 Glc- cells display a normal pattern of N-glycan processing in both growth phases. We therefore conclude that N-glycan processing may be used as an early biochemical marker of the enterocytic differentiation process of HT-29 cells.     

fMRI evidence for a cortical hierarchy of pitch pattern processing

Pitch patterns, such as melodies, consist of two levels of structure: a global level, comprising the pattern of ups and downs, or contour; and a local level, comprising the precise intervals that make up this contour. An influential neuropsychological model suggests that these two levels of processing are hierarchically linked, with processing of the global structure occurring within the right hemisphere in advance of local processing within the left. However, the predictions of this model and its anatomical basis have not been tested in neurologically normal individuals. The present study used fMRI and required participants to listen to consecutive pitch sequences while performing a same/different one-back task. Sequences, when different, either preserved (local) or violated (global) the contour of the sequence preceding them. When the activations for the local and global conditions were contrasted directly, additional activation was seen for local processing in right planum temporale and posterior superior temporal sulcus (pSTS). The presence of additional activation for local over global processing supports the hierarchical view that the global structure of a pitch sequence acts as a "framework" on which the local detail is subsequently hung. However, the lateralisation of activation seen in the present study, with global processing occurring in left pSTS and local processing occurring bilaterally, differed from that predicted by the neuroanatomical model. A re-examination of the individual lesion data on which the neuroanatomical model is based revealed that the lesion data equally well support the laterality scheme suggested by our data. While the present study supports the hierarchical view of local and global processing, there is an evident need for further research, both in patients and neurologically normal individuals, before an understanding of the functional lateralisation of local and global processing can be considered established.     

Interpreting Quantifier Scope Ambiguity: Evidence of Heuristic First,Algorithmic Second Processing

The present work suggests that sentence processing requires both heuristic and algorithmic processing streams, where the heuristic processing strategy precedes the algorithmic phase. This conclusion is based on three self-paced reading experiments in which the processing of two-sentence discourses was investigated, where context sentences exhibited quantifier scope ambiguity. Experiment 1 demonstrates that such sentences are processed in a shallow manner. Experiment 2 uses the same stimuli as Experiment 1 but adds questions to ensure deeper processing. Results indicate that reading times are consistent with a lexical-pragmatic interpretation of number associated with context sentences, but responses to questions are consistent with the algorithmic computation of quantifier scope. Experiment 3 shows the same pattern of results as Experiment 2, despite using stimuli with different lexical-pragmatic biases. These effects suggest that language processing can be superficial, and that deeper processing, which is sensitive to structure, only occurs if required. Implications for recent studies of quantifier scope ambiguity are discussed.     

Prostatic tissue protein alterations due to delayed time to freezing

Clinical studies often produce fresh tissue samples, which ideally should be immediately snap frozen for storage and subsequent analysis. However, this is often not practically possible, and there is inevitably a time period during which the sample is stored on ice. The delay in freezing may allow endogenous degradation of proteins to occur, affecting 2-D gel protein profiles. This study aims to investigate the type and extent of this degradation by examining how the time-to-freezing delay alters prostatic tissue protein profile. The prostate carcinoma-3 cell line (PC-3), prostate cancer xenografts and canine prostate were used with fluorescence 2-D DIGE to assess protein degradation. It was found that 30-min processing time had minimal effects on the protein profile. Longer delays had little visible effect, but subtle alterations in protein profile began to accumulate as time increased. These data support the practice of completing tissue processing as rapidly as possible, and indicate that short processing times do not notably perturb the 2-D gel spot pattern from prostatic tissue.     

一种感知视觉运动信息的简化脑模型

视觉运动信息的感知过程,包括从局域运动检测到对模式整体运动的感知过程.我们以蝇视觉系统的图形-背景相对运动分辨的神经回路网络为基本框架,采用初级运动检测器的六角形阵列作为输入层,构造了一种感知视觉运动信息的简化脑模型,模拟了运动信息应该神经计算模型各个层次上的处理.该模型对差分行为实验结果作出了正确预测.本文并对空间生理整合的神经机制作了讨论.     

Mueller矩阵在生物医学检测方面应用的实验研究

Mueller矩阵是公认的能很好地表述介质偏振特性的一种方法,由于散射光偏振在生物组织无创伤诊断技术等诸多领域中的重要应用价值,对组织散射特性的Mueller矩阵的研究成为国际上组织光学的热点之一。研究设计了一种新的测量Mueller矩阵的实验装置:斜入射正接收装置,并推导出一组后续数据处理的算法。由此所获得的Mueller矩阵空间分布图的清晰度不亚于其它所报道的,并且测量方法具有结构简单、方便、准确等优点。实验结果表明:入射角影响Mueller矩阵空间分布图;随着介质浓度的增大,随机介质后向散射Mueller矩阵各元素的空间分布图样减小;同时列举了真实生物组织样品(肌肉组织)的后向散射Mueller矩阵的实测结果,由此证明各向异性生物组织的后向散射Mueller矩阵各元素的空间分布图样与纤维的走向有关。     

Stimulus saliency modulates pre-attentive processing speed in human visual cortex

The notion of a saliency-based processing architecture [1] underlying human vision is central to a number of current theories of visual selective attention [e.g., 2]. On this view, focal-attention is guided by an overall-saliency map of the scene, which integrates (sums) signals from pre-attentive sensory feature-contrast computations (e.g., for color, motion, etc.). By linking the Posterior Contralateral Negativity (PCN) component to reaction time (RT) performance, we tested one specific prediction of such salience summation models: expedited shifts of focal-attention to targets with low, as compared to high, target-distracter similarity. For two feature-dimensions (color and orientation), we observed decreasing RTs with increasing target saliency. Importantly, this pattern was systematically mirrored by the timing, as well as amplitude, of the PCN. This pattern demonstrates that visual saliency is a key determinant of the time it takes for focal-attention to be engaged onto the target item, even when it is just a feature singleton.