Category learning induces position invariance of pattern recognition across the visual field

Human object recognition is considered to be largely invariant to translation across the visual field. However, the origin of this invariance to positional changes has remained elusive, since numerous studies found that the ability to discriminate between visual patterns develops in a largely location-specific manner, with only a limited transfer to novel visual field positions. In order to reconcile these contradicting observations, we traced the acquisition of categories of unfamiliar grey-level patterns within an interleaved learning and testing paradigm that involved either the same or different retinal locations. Our results show that position invariance is an emergent property of category learning. Pattern categories acquired over several hours at a fixed location in either the peripheral or central visual field gradually become accessible at new locations without any position-specific feedback. Furthermore, categories of novel patterns presented in the left hemifield are distinctly faster learnt and better generalized to other locations than those learnt in the right hemifield. Our results suggest that during learning initially position-specific representations of categories based on spatial pattern structure become encoded in a relational, position-invariant format. Such representational shifts may provide a generic mechanism to achieve perceptual invariance in object recognition.

Measuring the spatial arrangement patterns of pathological lesions in histological sections of brain tissue

The development of abnormal protein aggregates in the form of extracellular plaques and intracellular inclusions is a characteristic feature of many neurodegenerative diseases such as Alzheimer's disease (AD), Creutzfeldt-Jakob disease (CJD) and the fronto-temporal dementias (FTD). An important aspect of a pathological protein aggregate is its spatial topography in the tissue. Lesions may not be randomly distributed within a histological section but exhibit spatial pattern, a departure from randomness either towards regularity or clustering. Information on the spatial pattern of a lesion may be useful in elucidating its pathogenesis and in studying the relationships between different lesions. This article reviews the methods that have been used to study the spatial topography of lesions. These include simple tests of whether the distribution of a lesion departs significantly from random using randomized points or sample fields, and more complex methods that employ grids or transects of contiguous fields and which can detect the intensity of aggregation and the sizes, distribution and spacing of the clusters. The usefulness of these methods in elucidating the pathogenesis of protein aggregates in neurodegenerative disease is discussed.

What does the study of the spatial patterns of pathological lesions tell us about the pathogenesis of neurodegenerative disorders?

Discrete pathological lesions, which include extracellular protein deposits, intracellular inclusions and changes in cell morphology, occur in the brain in the majority of neurodegenerative disorders. These lesions are not randomly distributed in the brain but exhibit a spatial pattern, that is, a departure from randomness towards regularity or clustering. The spatial pattern of a lesion may reflect pathological processes affecting particular neuroanatomical structures and, therefore, studies of spatial pattern may help to elucidate the pathogenesis of a lesion and of the disorders themselves. The present article reviews first, the statistical methods used to detect spatial patterns and second, the types of spatial patterns exhibited by pathological lesions in a variety of disorders which include Alzheimer's disease, Down syndrome, dementia with Lewy bodies, Creutzfeldt-Jakob disease, Pick's disease and corticobasal degeneration. These studies suggest that despite the morphological and molecular diversity of brain lesions, they often exhibit a common type of spatial pattern (i.e. aggregation into clusters that are regularly distributed in the tissue). The pathogenic implications of spatial pattern analysis are discussed with reference to the individual disorders and to studies of neurodegeneration as a whole.

Differences in the density and spatial distribution of florid and diffuse plaques in variant Creutzfeldt-Jakob disease (vCJD)

Objective: To determine whether in cases of variant Creutzfeldt-Jakob disease (vCJD), the florid-type plaques are derived from the diffuse plaques or whether the 2 plaque types develop independently. Material: Blocks of frontal, parietal, occipital and temporal neocortex and cerebellar cortex from 11 cases of vCJD. Method: The density, distribution and spatial pattern of the florid and diffuse plaques were determined in each brain region using spatial pattern analysis. Results: The density of the diffuse plaques was significantly greater than that of the florid plaques in most areas. The ratio of the diffuse to florid plaques varied between brain regions and was maximal in the molecular layer of the cerebellum. The densities of the florid and diffuse plaques were positively correlated in the parietal cortex, occipital cortex, the inferior temporal gyrus and the dentate gyrus. Plaque densities were not related to disease duration. In the cerebral cortex, the diffuse plaques were more commonly evenly distributed or occurred in large clusters along the cortex parallel to the pia mater compared with the florid plaques which occurred more frequently in regularly distributed clusters. Conclusion: The florid plaques may not be derived from the diffuse plaques, the 2 plaque types appearing to develop independently with unique factors involved in their pathogenesis.

The spatial patterns of prion protein deposits in cases of variant Creutzfeldt-Jakob disease

The spatial patterns of the prion protein (PrP) deposits were studied in immunostained sections of areas of the cerebral cortex, hippocampus, dentate gyrus, and the molecular layer of the cerebellum in 11 cases of variant Creutzfeldt-Jakob disease (vCJD). Clustering of PrP deposits, with a regular distribution of the clusters parallel to the tissue boundary, was the most common spatial pattern observed. Two morphological types of PrP deposit were recognised, those consisting of a condensed core (florid deposits) and those deposits lacking a condensed core (non-florid deposits). The florid and non-florid PrP deposits exhibited a different profile of spatial patterns. First, the florid deposits exhibited a regularly distributed pattern of clusters more frequently than the non-florid deposits. Second, the florid deposits formed larger clusters (greater than1,600 µm in diameter) less frequently than the non-florid deposits. In the areas of the cerebral cortex that exhibited a regular distribution of PrP deposit clusters, the cluster size of the deposits approximated that of the groups of cells of the cortico-cortical pathway origin in only 12% of analyses. No significant differences in the frequency of the different types of spatial pattern were observed in different brain regions, or in the cerebral cortex between the upper and lower laminae. It was concluded that the spatial patterns of the PrP deposits in the cerebral cortex in vCJD are unlikely to reflect the degeneration of the cortico-cortical pathways as has been reported in sporadic CJD (sCJD). In addition, different factors could be involved in the development of the deposits with and without a condensed core.

Spatial patterns of the vacuolation in subcortical white matter in sporadic Creutzfeldt-Jakob disease (sCJD)

OBJECTIVE: To study the spatial patterns of the vacuolation ("spongiform change") in the subcortical white matter in the "classical" form of sporadic Creutzfeldt-Jakob disease (sCJD). MATERIAL: Frontal, parietal, occipital and temporal lobes of 11 cases of sCJD. METHOD: Spatial patterns were studied across the white matter at the base of the gyri using spatial pattern analysis. RESULTS: In the white matter of all gyri studied, vacuoles were aggregated into clusters, 50 to > 800 microm in diameter and in 22/37 (59%) of gyri, the clusters of vacuoles exhibited a regular distribution across the base of the gyri. In the remaining gyri, the vacuoles were aggregated into large clusters, at least 400 microm or 800 microm in diameter, but without evidence of a regular distribution. In a significant proportion of gyri, the spatial patterns of the vacuolation were similar to those reported previously for spongiform change and prion protein (PrP) deposits in the corresponding grey matter. CONCLUSIONS: Degeneration of the white matter and the formation of clusters of vacuoles may occur before the degeneration of the grey matter or could be a consequence of pathology affecting the cortico-cortical pathways.

Spatial patterns of β-amyloid (Aβ) deposits in familial and sporadic Alzheimer's disease

The spatial patterns of the diffuse, primitive, and classic β-amyloid (Aβ) deposits were compared in cortical regions in early-onset familial Alzheimer's disease (EO-FAD) linked to mutations of the amyloid precursor protein APP) or presenilin 1 (PSEN1) genes, late-onset familial AD (LO-FAD), and sporadic AD (SAD). The objective was to determine whether genetic factors influenced the spatial patterns of the Aβ deposits. Aβ deposits were distributed either in clusters which were regularly distributed parallel to the pia mater or in larger, non-regularly distributed clusters. There were no significant differences in spatial pattern of the diffuse deposits between patient groups but mean cluster size of the diffuse deposits was larger in FAD compared with SAD. Primitive Aβ deposits were more frequently distributed in regular clusters and less frequently distributed in large clusters in FAD compared with SAD. Classic Aβ deposits were more frequently distributed in regularly spaced clusters and less frequently distributed in large clusters in LO-FAD compared with EO-FAD. There were no significant differences in the spatial patterns or cluster sizes of Aβ deposits in cases classified according to apolipoprotein E (APOE) genotype. These results suggest (1) greater deposition of Aβ in the form of clusters of diffuse deposits in FAD, (2) a greater proportion of diffuse deposits may be converted to primitive deposits in SAD, (3) classic deposits are more widely distributed in EO-FAD, and (4) the presence of APOE allele ε4 has little effect on the spatial patterns of Aβ deposits.

The spatial patterns of beta-amyloid deposits and neurofibrillary tangles in the cerebral cortex in Alzheimer's disease

In Alzheimer's disease (AD) brain, beta-amyloid (Abeta) deposits and neurofibrillary tangles (NFT) are not randomly distributed but exhibit a spatial pattern, i.e., a departure from randomness towards regularity or clustering. Studies of the spatial pattern of a lesion may contribute to an understanding of its pathogenesis and therefore, of AD itself. This article describes the statistical methods most commonly used to detect the spatial patterns of brain lesions and the types of spatial patterns exhibited by ß-amyloid deposits and NFT in the cerebral cortex in AD. These studies suggest that within the cerebral cortex, Abeta deposits and NFT exhibit a similar spatial pattern, i.e., an aggregation of individual lesions into clusters which are regularly distributed parallel to the pia mater. The location, size and distribution of these clusters supports the hypothesis that AD is a 'disconnection syndrome' in which degeneration of specific cortical pathways results in the formation of clusters of NFT and Abeta deposits. In addition, a model to explain the development of the pathology within the cerebral cortex is proposed.

Analysis of spatial patterns in histological sections of brain tissue using a method based on regression

A method of determining the spatial pattern of any histological feature in sections of brain tissue which can be measured quantitatively is described and compared with a previously described method. A measurement of a histological feature such as density, area, amount or load is obtained for a series of contiguous sample fields. The regression coefficient (β) is calculated from the measurements taken in pairs, first in pairs of adjacent samples and then in pairs of samples taken at increasing degrees of separation between them, i.e. separated by 2, 3, 4,..., n units. A plot of β versus the degree of separation between the pairs of sample fields reveals whether the histological feature is distributed randomly, uniformly or in clusters. If the feature is clustered, the analysis determines whether the clusters are randomly or regularly distributed, the mean size of the clusters and the spacing of the clusters. The method is simple to apply and interpret and is illustrated using simulated data and studies of the spatial patterns of blood vessels in the cerebral cortex of normal brain, the degree of vacuolation of the cortex in patients with Creutzfeldt-Jacob disease (CJD) and the characteristic lesions present in Alzheimer's disease (AD). Copyright (C) 2000 Elsevier Science B.V.

Analysis of spatial patterns in histological sections of brain tissue

A method is described which enables the spatial pattern of discrete objects in histological sections of brain tissue to be determined. The method can be applied to cell bodies, sections of blood vessels or the characteristic lesions which develop in the brain of patients with neurodegenerative disorders. The density of the histological feature under study is measured in a series of contiguous sample fields arranged in a grid or transect. Data from adjacent sample fields are added together to provide density data for larger field sizes. A plot of the variance/mean ratio (V/M) of the data versus field size reveals whether the objects are distributed randomly, uniformly or in clusters. If the objects are clustered, the analysis determines whether the clusters are randomly or regularly distributed and the mean size of the clusters. In addition, if two different histological features are clustered, the analysis can determine whether their clusters are in phase, out of phase or unrelated to each other. To illustrate the method, the spatial patterns of senile plaques and neurofibrillary tangles were studied in histological sections of brain tissue from patients with Alzheimer's disease.

The spatial patterns of β/A4 deposit subtypes in Down's syndrome

The spatial patterns of diffuse, primitive and classic β/A4 deposits were studied in coronal sections of the hippocampus and adjacent gyri in 11 cases of Down's syndrome (DS) varying in age from 38 to 67 years. The objectives of the study were first, to compare the spatial patterns of β/A4 deposits revealed in DS with those reported in cases of Alzheimer's disease (AD) and second, to study how the spatial patterns of β/A4 deposits may develop in the tissue. The spatial patterns revealed in DS exhibited a number of similarities with those reported in AD: (1) the range and frequency of the different types of spatial pattern revealed were similar, (2) β/A4 deposits occurred in clusters and in many cortical tissues, the clusters were distributed in a regular pattern parallel to the pia, (3) the clusters of diffuse and primitive β/A4 deposits occurred in an alternating pattern along the cortex, and (4) the clusters of classic β/A4 deposits were not correlated with the clusters of the diffuse and primitive deposits. Primitive deposits may develop from the diffuse deposits in regions of the cortex where extracellular paired helical filaments were formed. However, clusters of the classic β/A4 deposits, which are formed in older cases, appear to develop independently of the diffuse and primitive deposits. © 1994 Springer-Verlag.

A morphometric study of the spatial patterns of TDP-43 immunoreactive neuronal inclusions in frontotemporal lobar degeneration (FTLD) with progranulin (GRN) mutation

Mutations of the progranulin (GRN) gene are a major cause of familial frontotemporal lobar degeneration with transactive response (TAR) DNA-binding protein of 43 kDa (TDP-43) proteinopathy (FTLD-TDP). We studied the spatial patterns of TDP-43 immunoreactive neuronal cytoplasmic inclusions (NCI) and neuronal intranuclear inclusions (NII) in histological sections of the frontal and temporal lobe in eight cases of FTLD-TDP with GRN mutation using morphometric methods and spatial pattern analysis. In neocortical regions, the NCI were clustered and the clusters were regularly distributed parallel to the pia mater; 58% of regions analysed exhibiting this pattern. The NII were present in regularly distributed clusters in 35% of regions but also randomly distributed in many areas. In neocortical regions, the sizes of the regular clusters of NCI and NII were 400-800 µm, approximating to the size of the modular columns of the cortico-cortical projections, in 31% and 36% of regions respectively. The NCI and NII also exhibited regularly spaced clustering in sectors CA1/2 of the hippocampus and in the dentate gyrus. The clusters of NCI and NII were not spatially correlated. The data suggest degeneration of the cortico-cortical and cortico-hippocampal pathways in FTLD-TDP with GRN mutation, the NCI and NII affecting different clusters of neurons.

Pooling and suppression in human spatial vision

A distinct feature of several recent models of contrast masking is that detecting mechanisms are divisively inhibited by a broadly tuned ‘gain pool’ of narrow-band spatial pattern mechanisms. The contrast gain control provided by this ‘cross-channel’ architecture achieves contrast normalisation of early pattern mechanisms, which is important for keeping them within the non-saturating part of their biological operating characteristic. These models superseded earlier ‘within-channel’ models, which had supposed that masking arose from direct stimulation of the detecting mechanism by the mask. To reveal the extent of masking, I measured the levels produced with large ranges of pattern spatial relationships that have not been explored before. Substantial interactions between channels tuned to different orientations and spatial frequencies were found. Differences in the masking levels produced with single and multiple component mask patterns provided insights into the summation rules within the gain pool. A widely used cross-channel masking model was tested on these data and was found to perform poorly. The model was developed and a version in which linear summation was allowed between all components within the gain pool but with the exception of the self-suppressing route typically provided the best account of the data. Subsequently, an adaptation paradigm was used to probe the processes underlying pooled responses in masking. This delivered less insight into the pooling than the other studies and areas were identified that require investigation for a new unifying model of masking and adaptation. In further experiments, levels of cross-channel masking were found to be greatly influenced by the spatio-temporal tuning of the channels involved. Old masking experiments and ideas relying on within-channel models were re-elevated in terms of contemporary cross-channel models (e.g. estimations of channel bandwidths from orientation masking functions) and this led to different conclusions than those originally arrived at. The investigation of effects with spatio-temporally superimposed patterns is focussed upon throughout this work, though it is shown how these enquiries might be extended to investigate effects across spatial and temporal position.

Spatial patterns of TDP-43 neuronal cytoplasmic inclusions (NCI) in fifteen cases of frontotemporal lobar degeneration with TDP-43 proteinopathy (FTLD-TDP)

Neuronal cytoplasmic inclusions (NCI) immunoreactive for transactive response DNA-binding protein (TDP-43) are the pathological hallmark of frontotemporal lobar degeneration with TDP-43 proteinopathy (FTLD-TDP). We studied the spatial patterns of the TDP-43 immunoreactive NCI in the frontal and temporal cortex of 15 cases of FTLD-TDP. The NCI were distributed parallel to the tissue boundary predominantly in regular clusters 50-400 µm in diameter. In five cortical areas, the size of the clusters approximated to the cells of the cortico-cortical pathways. In most regions, cluster size was smaller than 400 µm. There were no significant differences in spatial patterns between familial and sporadic cases. Cluster size of the NCI was not correlated with disease duration, brain weight, Braak stage, or disease subtype. The spatial pattern of the NCI was similar to that of neuronal inclusions in other neurodegenerative diseases and may reflect a common pattern of degeneration involving the cortico-cortical projections.

Different molecular pathologies result in similar spatial patterns of cellular inclusions in neurodegenerative disease:a comparative study of eight disorders

Recent research suggests cell-to-cell transfer of pathogenic proteins such as tau and α-synuclein may play a role in neurodegeneration. Pathogenic spread along neural pathways may give rise to specific spatial patterns of the neuronal cytoplasmic inclusions (NCI) characteristic of these disorders. Hence, the spatial patterns of NCI were compared in four tauopathies, viz., Alzheimer's disease, Pick's disease, corticobasal degeneration, and progressive supranuclear palsy, two synucleinopathies, viz., dementia with Lewy bodies and multiple system atrophy, the 'fused in sarcoma' (FUS)-immunoreactive inclusions in neuronal intermediate filament inclusion disease, and the transactive response DNA-binding protein (TDP-43)-immunoreactive inclusions in frontotemporal lobar degeneration, a TDP-43 proteinopathy (FTLD-TDP). Regardless of molecular group or morphology, NCI were most frequently aggregated into clusters, the clusters being regularly distributed parallel to the pia mater. In a significant proportion of regions, the regularly distributed clusters were in the size range 400-800 μm, approximating to the dimension of cell columns associated with the cortico-cortical pathways. The data suggest that cortical NCI in different disorders exhibit a similar spatial pattern in the cortex consistent with pathogenic spread along anatomical pathways. Hence, treatments designed to protect the cortex from neurodegeneration may be applicable across several different disorders. © 2012 Springer-Verlag.