Finkel, Leif H
Email Address
ORCID
Disciplines
Search Results
Now showing 1 - 7 of 7
Publication Principal Component Analysis of Temporal and Spatial Information for Human Gait Recognition(2005-09-01) Das, Sandhitsu; Lazarewicz, Maciej; Finkel, Leif HPrincipal component analysis was applied to human gait patterns to investigate the role and relative importance of temporal versus spatial features. Datasets consisted of various limb and body angles sampled over increasingly long time intervals. We find that spatial and temporal cues may be useful for different aspects of recognition. Temporal cues contain information that can distinguish the phase of the gait cycle; spatial cues are useful for distinguishing running from walking. PCA and related techniques may be useful for identifying features used by the visual system for recognizing biological motion.Publication A Computational Model of the Nucleus Accumbens: Network Properties and their Functional Implications(2003-03-20) Wolf, John A.; Finkel, Leif HThe Nucleus accumbens integrates convergent input from a number of limbic structures, and has been implicated in a variety of behavioral disorders including addiction and schizophrenia. The bistable membrane properties of the principal cell in the NAcb, the GABAergic medium spiny projection neuron (MSP), have been proposed to mediate afferent integration. To investigate how intrinsic properties may underlie this mechanism, we constructed a model of an MSP neuron in GENESIS, which preserves the main morphological features and relevant ionichynaptic currents. The model captures the major properties of in vivo neurons, including a non-linear response to the number of afferent inputs. In order to examine network properties of the NAcb and its response to varying patterns of afferent input, a 100- cell network with modifiable levels of gap junctions and GABAergic synaptic connectivity was constructed. Afferent inputs were modeled as Poisson-distributed spike trains. Addition of lateral inhibition in the network led to a decrease in spike output for cells receiving less synchronized input, suggesting that this may be a mechanism for increasing the signal to noise ratio. Dopaminergic modulation of the whole network led to a slight increase in overall synchronization, but did not further segregate cells that were already receiving synchronous input.Publication Afferent Stream Integration in a Model of the Nucleus Accumbens(2001-10-25) Wolf, John A.; Contreras, Diego; Finkel, Leif H; Schroeder, Lee F.The Nucleus Accumbens (Nacc) receives convergent input from a number of structures including prefrontal cortex, hippocampus, and amygdala, as well as substantial dopaminergic modulation. The principle cell in the Nacc, the medium spiny projection neuron (MSP), has bistable properties that have been proposed as a primary mediator of the integrative properties in the Nacc. In order to determine the minimum biophysical properties required to generate a nonlinear "bistable" membrane potential, we constructed a 29 compartment MSP cell in GENESIS. This included an inward rectifying K+ (KIR), an A-current K+, HVA Ca2+ and suitable fast Na+ and delayed rectifier K+ channels. Various amounts and distributions of afferent input were examined to determine the amount of coincident input required to move the membrane from the "down" state (-85 mV) to the "up" state (-60 mV) and to fire the cell. The number of synchronous afferents required was substantially higher than previously estimated. We modeled the effect of dopaminergic modulation by increasing the conductance of the KIR and Ca2+ channels, demonstrating that the response to input is dependent on state. In parallel studies, in vivo extracellular recordings were obtained from Nacc and neocortex in anesthetized mice. Simultaneous recordings revealed that cortex and Nacc oscillate in synchrony. Lesions that disrupt slow oscillations in the cortex alter rhythmicity in the Nacc, suggesting interdependency of the mechanisms generating bistability.Publication The Role of NMDA Currents in State Transitions of the Nucleus Accumbens Medium Spiny Neuron(2005-06-01) Wolf, John A.; Moyer, Jason T; Finkel, Leif H.The nucleus accumbens (NAcb) integrates information from a wide range of glutamatergic afferent inputs, including the prefrontal cortex, hippocampus and amygdala. One of the glutamatergic receptors, the NMDA channel, has been implicated in the non-linearity of the current–voltage relationship in these cells under certain input conditions. In order to examine the relationship of the different glutamatergic receptors to the membrane response, we modeled the AMPA, GABAA and NMDA receptors in the medium spiny (MSP) cells and their afferent input. The model demonstrates that the NMDA current is capable of sustaining certain membrane states and contributes to the non-linearity of the membrane response to input.Publication Curvature Covariation as a Factor in Perceptual Salience(2003-03-20) Murphy, Thomas M.; Matlin, Mark; Finkel, Leif HThe salience of a contour depends upon several factors, including continuity, closure and curvature consistency. We analyze the statistics of curvature variation using a single image from Shimon Ullman’s [1] original work on contour salience. We develop a measure based on the arc length of a contour segment over which curvature variation remains within a constrained range. Locally, all contours in the image are similar with respect to curvature consistency. However, when the entire contour is considered, the most salient contours are found to have the most consistent curvatures. This finding reinforces Ullman’s point that salience is a global property of the object. We interpret these results in view of Rosenholtz’s [2] recent model of salience as a statistical measure of outliers from a population. In addition, we speculate on the visual cortical mechanisms in striate and extrastriate cortex required to carry out salience measurements on this class of images.Publication Computational Modeling of Medium Spiny Projection Neurons in Nucleus Accumbens: Toward the Cellular Mechanisms of Afferent Stream Integration(2001-07-01) Wolf, John A.; Schroeder, Lee F.; Finkel, Leif HThe nucleus accumbens (Nacc) regulates the major feedback pathways linking prefrontal cortex, hippocampus, and amygdala. We describe simulations of a biophysical level model of a single medium spiny projection (MSP) neuron, the principle cell of the Nacc. The model suggests that the unusual bistable membrane potential of MSP cells arises from the interplay between two potassium currents, KIR and KA. We find that the transition from the membrane potential down state (~-85mV) to the upstate (~-60mV)requires a significant barrage of synchronized inputs, and that ongoing afferent stimulation is required to maintain the cell in the up state. The Nacc receives the densest dopaminergic innervation in the brain, and the model demonstrates, in agreement with recent experimental evidence, that dopamine acts to increase the energy barrier to membrane potential state transitions. Through its action on KIR and L-type Ca2+ channels, dopamine selectively lowers cell gain in the down state and increases it in the up state, a mechanism for context-dependent gain control. These findings suggest a mechanism of afferent pattern integration in the accumbens arising from transient synchronization among ensembles of MSP neurons. We attempt to relate these findings to possible origins of abnormalities of sensory gating in schizophrenia.Publication Hyperspectral Image Processing Using Locally Linear Embedding(2003-03-20) Kim, David H.; Finkel, Leif HWe describe a method of processing hyperspectral images of natural scenes that uses a combination of k-means clustering and locally linear embedding (LLE). The primary goal is to assist anomaly detection by preserving spectral uniqueness among the pixels. In order to reduce redundancy among the pixels, adjacent pixels which are spectrally similar are grouped using the k-means clustering algorithm. Representative pixels from each cluster are chosen and passed to the LLE algorithm, where the high dimensional spectral vectors are encoded by a low dimensional mapping. Finally, monochromatic and tri-chromatic images are constructed from the k-means cluster assignments and LLE vector mappings. The method generates images where differences in the original spectra are reflected in differences in the output vector assignments. An additional benefit of mapping to a lower dimensional space is reduced data size. When spectral irregularities are added to a patch of the hyperspectral images, again the method successfully generated color assignments that detected the changes in the spectra.