Ralph M. Siegel

Pattern recognition of the electroencephalogram by artificial neural networks

A back-propagation network was trained to recognize high voltage spike-and-wave spindle (HVS) patterns in the rat, a rodent model of human petit mal epilepsy. The spontaneously occurring HVSs were examined in 137 rats of the Fisher 344 and Brown Norway strains and their F1, F2 and backcross hybrids. Neocortical EEG and movement of the rat were recorded for 12 night hours in each animal and analog data were filtered (low cut: 1 Hz; high cut: 50 Hz) and sampled at 100 Hz with 12 bit precision. A training data set was generated by manually marking durations of HVS epochs in 16 representative animals selected from each group. Training data were presented to back-propagation networks with variable numbers of input, hidden and output cells. The performance of different types of networks was first examined with the training samples and then the best configuration was tested on novel sets of the EEG data. FFT transformation of EEG significantly improved the pattern recognition ability of the network. With the most effective configuration (16 input; 19 hidden; 1 output cells) the summed squared error dropped by 80% as compared with that of the initial random weights. When testing the network with new patterns the manual and automatic evaluations were compared quantitatively. HVSs which were detected properly by the network reached 93-99% of the manually marked HVS patterns, while falsely detected events (non-HVS, artifacts) varied between 18% and 40%. These findings demonstrate the utility of back-propagation networks in automatic recognition of EEG patterns.

Optic Flow Selectivity in the Anterior Superior Temporal Polysensory Area, STPa, of the Behaving Monkey

Earlier studies of neurons in the anterior region of the superior temporal polysensory area (STPa) have demonstrated selectivity for visual motion using stimuli contaminated by nonmotion cues, including texture, luminance, and form. The present experiments investigated the motion selectivity of neurons in STPa in the absence of form cues using random dot optic flow displays. The responses of neurons were tested with translation, rotation, radial, and spiral optic flow displays designed to mimic the types of motion that occur during locomotion. Over half of the neurons tested responded significantly to at least one of these displays. On a cell by cell basis, 60% of the neurons tested responded selectively to rotation, radial, and spiral motion, whereas 20% responded selectively to translation motion. The majority of neurons responded maximally to single-component optic flow displays but was also significantly activated by the spiral displays that contained their preferred component. Moreover, there was a bias in the selectivity of the neurons for radial expansion motion. These results suggest that neurons within STPa are contributing to the analysis of optic flow. Furthermore, the preponderance of cells selective for radial expansion provides evidence that this area may be specifically involved in the processing of forward locomotion and/or looming stimuli. Finally, these results provide carefully controlled physiological evidence for an extension and specialization of the motion-processing pathway into the anterior temporal lobe.

Functional Architecture of Spatial Attention in the Parietal Cortex of the Behaving Monkey

Functional architectures facilitate orderly transmittal of representations between cortices, allow for local interactions between neurons, and ensure a uniform distribution of feature representations with respect to larger-scale topographies. We sought to correlate such topographies with internal cognitive states. A psychophysical task for which the monkey was required to detect a change in one of two identical peripheral expanding flow fields tested for spatial shifts of attention. The monkey was cued as to which flow would change with a small cue near the fixation points. Reaction time data indicate that the monkeys performance in the optic flow detection task depended on the location of the cue. Using optical imaging of intrinsic signals, we show that a monkeys internally generated locus of attention is correlated with an 800-860 μm patchy topological architecture across the cortical surface of the inferior parietal lobule. The attentional patches vary in location but are stable in spatial frequency. The patches are embedded in a larger-scale and stable representation of eye position. Trial-by-trial analysis of the images indicates that the organizational scheme with simultaneous stable and variable subcomponents occurs within the experiment of 1 d, as well as across days. This novel functional architecture is the first to be correlated with attentional mechanisms and could support a fine-scale functional architecture underlying hemispatial neglect, an attentional deficit caused by parietal lesions.

Two-Photon Imaging of Calcium in Virally Transfected Striate Cortical Neurons of Behaving Monkey

Two-photon scanning microscopy has advanced our understanding of neural signaling in non-mammalian species and mammals. Various developments are needed to perform two-photon scanning microscopy over prolonged periods in non-human primates performing a behavioral task. In striate cortex in two macaque monkeys, cortical neurons were transfected with a genetically encoded fluorescent calcium sensor, memTNXL, using AAV1 as a viral vector. By constructing an extremely rigid and stable apparatus holding both the two-photon scanning microscope and the monkeys head, single neurons were imaged at high magnification for prolonged periods with minimal motion artifacts for up to ten months. Structural images of single neurons were obtained at high magnification. Changes in calcium during visual stimulation were measured as the monkeys performed a fixation task. Overall, functional responses and orientation tuning curves were obtained in 18.8% of the 234 labeled and imaged neurons. This demonstrated that the two-photon scanning microscopy can be successfully obtained in behaving primates.

Lack of selectivity for simple shapes defined by motion and luminance in Stpa of the behaving macaque

THE anterior superior temporal polysensory area (STPa) has been hypothesized to be an integration site for signals coming from the dorsal and ventral visual pathways. To determine whether neurons in STPa were selective for simple two-dimensional geometrical shapes and whether this area might integrate different visual cues, cells were tested with motion-and luminance-defined shapes. Many neurons were activated by the shape stimuli under at least one condition; however, very few showed selectivity for a particular shape under either condition. Only one neuron responded selectively to shapes defined by both cues. Thus selectivity for simple shapes is not a prevalent property of STPa neurons and the integration of luminance and motion signals does not appear to occur at the single neuron level in STPa.

Construction and Representation of Visual Space in the Inferior Parietal Lobule

In human subjects, neurological case histories have provided details of striking deficits of cognitive functions. These findings have been supplemented by comparative studies in nonhuman primates that permit invasive manipulations and measurements of brain tissue, as well as theories that provide a structure for the synthesis and distillation of such clinical and experimental results into general principles. The formation of visuospatial representations of the surrounding world has proven to be particularly amenable to this multifaceted approach. Although far from being completely understood, substantial progress has been made in the past 20 years in the analysis of this cognitive process. One patient, suffering from profound hemi-inattention has: lost her idea of ‘left,’ both with regard to the world and her own body. Sometimes (when eating) she complains that her portions are too small, but this is because she eats only from the right half of the plate—it does not occur to her that it has a left half as well. Sometimes, she will put on lipstick, and make tip the right half of her face, leaving the left half completely neglected: It is almost impossible to treat these things, because her attention cannot be drawn to them, and she has no conception that they are wrong. She knows it intellectually, and can understand, and laugh; but it is impossible for her to know it directly. (Sacks, 1985) This particular type of spatial and attentional deficit arises from damage to associational cortical regions which were initially defined on the basis of late myelination during development (Flechsig, 1876). This patient does not have a hemianopia (half-vision field blindness), for her vision is normal. But her cognitive disassociations are in a sense more profound than blunt dissections of primary sensory or motor function. Removal of primary sensory or motor cortex leads to devastating blindness, deafness, or paralysis. But such loss can be overcome, at least in the sensory modality, by reliance on spared sensory capabilities. However, damage to the association cortex strikes at the very core of being. The patient can see but cannot attend to half of her visual world. She cannot intellectually compensate, because her spatial intellect itself is damaged. The challenge to neuroscience is to generate mechanistic explanations, perhaps reductionist, perhaps synthetic, to account for normal cognitive deficits on this scale. These explanations might explain cognitive function in the normal individual.

A functional architecture of optic flow in the inferior parietal lobule of the behaving monkey.

The representation of navigational optic flow across the inferior parietal lobule was assessed using optical imaging of intrinsic signals in behaving monkeys. The exposed cortex, corresponding to the dorsal-most portion of areas 7a and dorsal prelunate (DP), was imaged in two hemispheres of two rhesus monkeys. The monkeys actively attended to changes in motion stimuli while fixating. Radial expansion and contraction, and rotation clockwise and counter-clockwise optic flow stimuli were presented concentric to the fixation point at two angles of gaze to assess the interrelationship between the eye position and optic flow signal. The cortical response depended upon the type of flow and was modulated by eye position. The optic flow selectivity was embedded in a patchy architecture within the gain field architecture. All four optic flow stimuli tested were represented in areas 7a and DP. The location of the patches varied across days. However the spatial periodicity of the patches remained constant across days at ∼950 and 1100 µm for the two animals examined. These optical recordings agree with previous electrophysiological studies of area 7a, and provide new evidence for flow selectivity in DP and a fine scale description of its cortical topography. That the functional architectures for optic flow can change over time was unexpected. These and earlier results also from inferior parietal lobule support the inclusion of both static and dynamic functional architectures that define association cortical areas and ultimately support complex cognitive function.

A decoding problem in dynamics and in number theory

Given a homeomorphism f of the circle, any splitting of this circle in two semiopen arcs induces a coding process for the orbits of f, which can be determined by recording the successive arcs visited by the orbit. The problem of describing these codes has a two hundred year history (that we briefly recall) in the particular case when the arcs are limited by a point and its image; in modern language, it is the kneading theory of such maps, and as such is relevant for our understanding of dynamical problems involving oscillations. This paper deals with questions attached to the general case, a problem considered by many mathematicians in the 50s and 60s in the case where f is a rotation, and which has recently found some applications in physiology. We show that, except for trivial cases, any code determines the rotation number, up to the orientation, of the homeomorphism which generates it. In the case the code is periodic, we can also determine whether or not it can be generated in this way. An equivalent problem in arithmetic consists of finding +/-p, knowing a collection of classes in Z/qZ of the form {m,m+p,.,m+(k-1)p}, where 2</=k</=q-2 and p and q are relatively prime. We describe this equivalence, and give simple solutions of the decoding problem both in the dynamical context and in the number theoretic context.

Representation of Visual Space in Area 7a Neurons Using the Center of Mass Equation

The firing rate of neurons in parietal area 7a of the behaving Rhesus monkey with its head fixed incorporates both visual and eye position information. This neural tuning is not in an ego-centered coordinate space. This physiological result was unexpected as behavioral deficits following parietal damage in human and monkey subjects suggested the existence of egocentric representations. A formulation to extract a world-centered system from area 7a neurons is presented that depends on the linearity of the eye position signal and the similarity of the equation describing the tuning of these neurons to the center of mass equation. This formulation permits the computation of the location of objects in world coordinates using either serial analysis of a single neurons activity or parallel processing of a collection of neurons. Experimental predictions are made for the relationship between different parameters of angle of gaze neurons.

Two-photon scanning microscopy of in vivo sensory responses of cortical neurons genetically encoded with a fluorescent voltage sensor in rat

A fluorescent voltage sensor protein “Flare” was created from a Kv1.4 potassium channel with YFP situated to report voltage-induced conformational changes in vivo. The RNA virus Sindbis introduced Flare into neurons in the binocular region of visual cortex in rat. Injection sites were selected based on intrinsic optical imaging. Expression of Flare occurred in the cell bodies and dendritic processes. Neurons imaged in vivo using two-photon scanning microscopy typically revealed the soma best, discernable against the background labeling of the neuropil. Somatic fluorescence changes were correlated with flashed visual stimuli; however, averaging was essential to observe these changes. This study demonstrates that the genetic modification of single neurons to express a fluorescent voltage sensor can be used to assess neuronal activity in vivo.