Modeling the Ventral and Dorsal Cortical Visual Pathways Using Artificial Neural Networks

Although in conventional models of cortical processing, object recognition and spatial properties are processed separately in ventral and dorsal cortical visual pathways respectively, some recent studies have shown that representations associated with both objects' identity (of shape) and space are present in both visual pathways. However, it is still unclear whether the presence of identity and spatial properties in both pathways have functional roles. In our study, we have tried to answer this question through computational modeling. Our simulation results show that both a model ventral and dorsal pathway, separately trained to do object and spatial recognition, respectively, each actively retained information about both identity and space. In addition, we show that these networks retained different amounts and kinds of identity and spatial information. As a result, our modeling suggests that two separate cortical visual pathways for identity and space (1) actively retain information about both identity and space (2) retain information about identity and space differently and (3) that this differently retained information about identity and space in the two pathways may be necessary to accurately and optimally recognize and localize objects. Further, modeling results suggests these findings are robust and do not strongly depend on the specific structures of the neural networks.

Lip movements enhance speech representations and effective connectivity in speech dorsal stream and its relationship with neurite architecture

Lip movements facilitate speech comprehension, especially under adverse listening conditions, but the neural mechanisms of this perceptual benefit at the phonemic and feature levels remain unclear. This fMRI study addresses this question by quantifying regional multivariate representation and network organization underlying audiovisual speech-in-noise perception. We found that valid lip movements enhanced neural representations of phoneme, place of articulation, or voicing feature of speech differentially in dorsal stream regions, including frontal speech motor areas and supramarginal gyrus. Such local changes were accompanied by strengthened dorsal stream effective connectivity. Moreover, the neurite orientation dispersion of left arcuate fasciculus, a structural basis of speech dorsal stream, predicted the visual enhancements of neural representations and effective connectivity. Our findings provide novel insight to speech science that lip movements promote both local phonemic and feature encoding and network connectivity in speech dorsal pathway and the functional enhancement is mediated by the microstructural architecture of the circuit.

Cross-Modal Interaction Between Auditory and Visual Input Impacts Memory Retrieval

How we perceive and learn about our environment is influenced by our prior experiences and existing representations of the world. Top-down cognitive processes, such as attention and expectations, can alter how we process sensory stimuli, both within a modality (e.g., effects of auditory experience on auditory perception), as well as across modalities (e.g., effects of visual feedback on sound localization). Here, we demonstrate that experience with different types of auditory input (spoken words vs. environmental sounds) modulates how humans remember concurrently-presented visual objects. Participants viewed a series of line drawings (e.g., picture of a cat) displayed in one of four quadrants while listening to a word or sound that was congruent (e.g., “cat” or <meow>), incongruent (e.g., “motorcycle” or <vroom–vroom>), or neutral (e.g., a meaningless pseudoword or a tonal beep) relative to the picture. Following the encoding phase, participants were presented with the original drawings plus new drawings and asked to indicate whether each one was “old” or “new.” If a drawing was designated as “old,” participants then reported where it had been displayed. We find that words and sounds both elicit more accurate memory for what objects were previously seen, but only congruent environmental sounds enhance memory for where objects were positioned – this, despite the fact that the auditory stimuli were not meaningful spatial cues of the objects’ locations on the screen. Given that during real-world listening conditions, environmental sounds, but not words, reliably originate from the location of their referents, listening to sounds may attune the visual dorsal pathway to facilitate attention and memory for objects’ locations. We propose that audio-visual associations in the environment and in our previous experience jointly contribute to visual memory, strengthening visual memory through exposure to auditory input.

Your head is there to move you around: Goal-driven models of the primate dorsal pathway

Neurons in the dorsal visual pathway of the mammalian brain are selective for motion stimuli, with the complexity of stimulus representations increasing along the hierarchy. This progression is similar to that of the ventral visual pathway, which is well characterized by artificial neural networks (ANNs) optimized for object recognition. In contrast, there are no image-computable models of the dorsal stream with comparable explanatory power. We hypothesized that the properties of dorsal stream neurons could be explained by a simple learning objective: the need for an organism to orient itself during self-motion. To test this hypothesis, we trained a 3D ResNet in a self-supervised manner to predict an agent's self-motion parameters from visual stimuli in a simulated environment. We found that the responses in this network accounted well for the selectivity of neurons in a large database of single-neuron recordings from the dorsal visual stream of non-human primates. In contrast, ANNs trained for action recognition or with a contrastive objective could not explain responses in the dorsal stream, despite also being trained on naturalistic videos with moving objects. These results demonstrate that an ecologically relevant, self-supervised cost function can account for dorsal stream properties in the primate brain.

Suppression of Luminance Contrast Sensitivity by Weak Color Presentation

The results of psychophysical studies suggest that color in a visual scene affects luminance contrast perception. In our brain imaging studies we have found evidence of an effect of chromatic information on luminance information. The dependency of saturation on brain activity in the visual cortices was measured by functional magnetic resonance imaging (fMRI) while the subjects were observing visual stimuli consisting of colored patches of various hues manipulated in saturation (Chroma value in the Munsell color system) on an achromatic background. The results indicate that the patches suppressed luminance driven brain activity. Furthermore, the suppression was stronger rather than weaker for patches with lower saturation colors, although suppression was absent when gray patches were presented instead of colored patches. We also measured brain activity while the subjects observed only the patches (on a uniformly black background) and confirmed that the colored patches alone did not give rise to differences in brain activity for different Chroma values. The chromatic information affects the luminance information in V1, since the effect was observed in early visual cortices (V2 and V3) and the ventral pathway (hV4), as well as in the dorsal pathway (V3A/B). In addition, we conducted a psychophysical experiment in which the ability to discriminate luminance contrast on a grating was measured. Discrimination was worse when weak (less saturated) colored patches were attached to the grating than when strong (saturated) colored patches or achromatic patches were attached. The results of both the fMRI and psychophysical experiments were consistent in that the effects of color were greater in the conditions with low saturation colors.

Unilateral Resection of Both Cortical Visual Pathways Alters Action but not Perception

The human cortical visual system consists of two major pathways, a ventral pathway that subserves perception and a dorsal pathway that subserves visuomotor control. These pathways follow dissociable developmental trajectories, and, accordingly, might be differentially susceptible to neurodevelopmental disorders or injuries. Previous studies have found that children with cortical resections of the ventral visual pathway retain largely normal visuoperceptual abilities. Whether visually guided actions, supported by computations carried out by the dorsal pathway, follow a similar pattern remains unknown. To address this question, we examined visuoperceptual and visuomotor behaviors in a pediatric patient, TC, who underwent a cortical resection that included portions of the left ventral and dorsal pathways. We collected data when TC used her right and left hands to perceptually estimate the width blocks that varied in width and length, and, separately, to grasp the same blocks. Her perceptual estimation performance was comparable to that of controls, independent of the hand used. In contrast, relative to controls, she showed reduced visuomotor sensitivity to object shape and this was more evident when she grasped the objects with her contralesional right hand. These results provide evidence for a striking difference in the reorganization profiles of the two visual pathways. This difference supports the notion that the two pathways exhibit differential susceptibility to neurodevelopmental disorders.

Somatosensory evoked potentials in Hirayama disease: A Brazilian study

Background: Hirayama’s disease (HD) is characterized by an insidious onset asymmetric weakness and atrophy of the forearm and hand. Taking as a premise, the etiopathogenesis of the disease is attributed to forward displacement of posterior wall of lower cervical dural canal in neck flexion causing marked compression and flattening of lower spinal cord. This may result in compression of the posterior column of the spinal cord and seems likely to result in somatosensory evoked potentials (SSEPs) abnormalities. In the present study, we studied the possible involvement of the lemniscal dorsal pathway in patients with HD. Methods: SSEPs in upper and lower extremities were prospectively performed in eight patients with HD. All the patients were recruited from the outpatient clinic of a neuromuscular disorder center from South Brazil. SSEPs were obtained by transcutaneous electrical stimulation of the median and posterior tibial nerves, on both sides. We collected the amplitude and the latency of the different components obtained in each channel. The interpretation was based on Brazilian study standards. Results: We evaluated seven men and one woman (mean age 27). The data obtained were compared to a control group consisting of eight patients with spondylotic cervical myelopathy, 6 men and 2 women with mean age of 59 years. The measurements of obtained by the SSEP were also compared between the groups and no significant difference was found for any of them. Conclusion: SSEP did not turn out to be an electrophysiological marker in our HD patients.