Plasticity and Learning

Talk Session: Sunday, May 19, 2024, 5:15 – 7:15 pm, Talk Room 1

Talk 1, 5:15 pm

Differential functional reorganization of ventral and dorsal visual pathways following childhood hemispherectomy

Vladislav Ayzenberg1 (), Michael Granovetter2, Sophia Robert2, Christina Patterson3, Marlene Behrmann 4; 1Department of Psychology, University of Pennsylvania, 2Department of Psychology, Carnegie Mellon University, 3School of Medicine, University of Pittsburgh, 4Department of Ophthalmology, University of Pittsburgh

A key signature of neural development is that the brains of children have a greater capacity for recovery following damage or surgery than adults. Indeed, children who have had an entire hemisphere of their brain removed to treat epilepsy (a procedure known as hemispherectomy) show a high degree of perceptual functioning despite the loss of both ventral and dorsal visual pathways in one hemisphere. Yet, accumulating evidence suggests that the dorsal pathway may mature earlier than the ventral pathway – raising the question of whether the two pathways also have a different capacity for recovery after surgery. In the current study, we sought to understand the extent to which functions of the ventral and dorsal pathways reorganize to the contralateral hemisphere following childhood hemispherectomy. We collected fMRI data from an equal number of left and right hemispherectomy patients (N = 8; age-at-surgery = 1-13 years; age-at-testing = 12-37 years) who completed tasks that typically elicit lateralized responses from the ventral or the dorsal pathway in controls, namely, word (left ventral), face (right ventral), tool (left dorsal), and global form (right dorsal) perception. Overall, there was greater evidence of functional reorganization in the ventral pathway than in the dorsal pathway. The majority of hemispherectomy patients showed normal degrees of word and face selectivity in their intact ventral pathway, despite losing the typically preferred hemisphere for each category. By contrast, only one patient showed evidence of normal selectivity for either tools or global form in their intact dorsal pathway. Importantly, because ventral and dorsal reorganization was tested within the same patients, these results cannot be explained by idiosyncratic factors such as disease etiology or age at the time of surgery. These findings suggest that the dorsal pathway has a shorter developmental window of plasticity than the ventral pathway because it matures earlier.

Talk 2, 5:30 pm

Patients with V1 damage exhibit increased orientation decoding in hMT+, but only if pulvinar is intact

Bing Li1 (), Tina T. Liu1, Matthew R. Cavanaugh2,3, Helena P. Bachmann1, Berkeley K. Fahrenthold2,3, Shruti Japee1, Krystel R. Huxlin2,3, Elisha P. Merriam1; 1Laboratory of Brain and Cognition, National Institute of Mental Health, NIH, Bethesda, MD, USA, 2Flaum Eye Institute, University of Rochester Medical Center, Rochester, NY, USA, 3Center for Visual Science, University of Rochester, Rochester, NY, USA

Orientation selectivity is a core property of V1 in mammals. Patients with V1 damage can relearn orientation discrimination at trained, blind-field locations. Here, we investigated a potentially key role for pathways bypassing V1, which directly transmit information to downstream visual cortical areas, generating orientation selectivity in these areas. We studied 3 stroke patients (33-63 y/o, all females): one with a large right V1 lesion, a second with a right V2/V3 lesion that spared V1, and a third with lesions affecting both right V1 and pulvinar. Participants were scanned with BOLD fMRI. They viewed small (2.5 deg radius), oriented (45 or 135 deg) gratings in the periphery (7.1-11.2 deg eccentricity) while performing a demanding task at fixation. Stimuli were presented either deep within the blind field or in a mirror-symmetric location in the intact hemifield. We also performed retinotopic mapping, scanning with an MT localizer, and T1-weighted structural scans. In the patient with extensive V1 damage, the ipsilesional hMT+ was visually responsive and able to decode orientation. However, in the patient with a V2/V3 lesion and the patient with right V1 plus pulvinar damage, the ipsilesional hMT+ was visually responsive, but orientation decoding failed to reach significance. Healthy controls exhibited significant orientation decoding in V1 but not in hMT+. Our findings suggest that after V1 damage, strong orientation selectivity may emerge in hMT+ from circuits bypassing V1, including via the pulvinar. When V1 is spared, even when V2/V3 are damaged, these alternative circuits do not generate robust orientation-selective BOLD signals in hMT+. Ongoing work is assessing the functional implications of these findings for perception and rehabilitation potential.

Acknowledgements: This work was supported by the Intramural Research Program of the National Institute of Mental Health (ZIAMH002966).

Talk 3, 5:45 pm

Training-Induced Functional Homogenization in the Occipitotemporal Cortex: Differential Cross-Modal Mechanisms in Blindness vs. Severe Low Vision

Lora Likova1 (), Zhangziyi Zhou1, Michael Liang1, Christopher Tyler1; 1Smith-Kettlewell Eye Research Institute

No visual motion is available to totally blind (TB) individuals, raising questions about the utilization of the territory of the visual motion complex (hMT+) in the absence of vision, and its potential reorganization through training. This study explores whether such reorganization differs between TB individuals and those with some residual vision, such as severe low vision (SLV). Methods: TB and SLV subjects underwent five sessions of the Cognitive-Kinesthetic Memory-Drawing Training for spatial navigation. Pre- and post-training, whole-brain scans (Siemens 3T Prisma) were conducted while subjects (i) haptically explored and memorized raised-line tactile maps (30 s); after a 20 s rest, they (ii) drew-from-haptic-memory (30 s) the maps using the opposite hand with a stylus. Results: Despite the absence of visual input, hMT+ was robustly activated bilaterally in TB individuals in the right-hand blind Memory-Drawing task. However, the left-hand Haptic Exploration task activated only right hMT+, an unexpected interhemispheric functional asymmetry. Furthermore, following the training, significant brain reorganization occurred in the lateral occipitotemporal cortex, forming clusters of cortical areas TPOJ 1-3, FST, MTG and LOd expressing the same task-response asymmetry as hMT+. In SLV individuals, on the other hand, although a similar large-scale functional clustering occurred, hMT+ was surprisingly excluded and even unilaterally suppressed. Granger Causal connectivity analysis revealed a complex interplay between hMT+, its surrounding cluster, and motor, somatosensory, and memory areas. Conclusions: The multidimensional findings shed light on non-visual hMT+ functionality, its novel interhemispheric asymmetries, and their implications for functional brain architecture and its reorganization through learning. Furthermore, the results reveal for the first time the emergence of training-induced functional-homogenization of extended clusters of occipitotemporal areas around hMT+ in the visually deprived, which in the sighted are functionally distinct, demonstrating a mechanism for a novel type of cross-modal functional reorganization, offering crucial insights into neuroplasticity and sensory compensation.

Acknowledgements: NIH/NEI EY024056 & NSF SL-CN1640914 to L. Likova

Talk 4, 6:00 pm

Individual differences of functional brain plasticity in central vision loss

Pinar Demirayak1 (), Rachel Chua1, Leland Fleming2, Kristina Visscher1; 1University of Alabama at Birmingham, Heersink School of Medicine, Department of Neurobiology, Birmingham, AL 35294, USA, 2McLean Hospital/Harvard Medical School, Boston, MA 02115, USA

Functional organization of the visual cortex is largely consistent across individuals. However, it is not clear to what extent the functional connectivity (FC) patterns between visual cortex and other areas vary across individuals, nor is it clear to what extent these connectivity patterns are shaped by experience. Central vision loss provides an excellent model to investigate visual system plasticity because different features of experience impact the same participants, i.e. sensory deprivation in central vision and increased usage of parts of peripheral vision. We studied whole-brain FC patterns for parts of primary visual cortex (V1) corresponding to parts of retina associated with increased and decreased use. We performed both group-level and individual-specific analyses in 21 participants with central vision loss and 22 participants with healthy or corrected-to-normal vision. Group-level FC results revealed that participants with central vision loss have reduced connectivity between sensory-deprived portions of V1 and temporo-parieto-occipital cortical areas, compared to controls. Group-level comparisons did not show any statistically significant difference in mean connection strength to areas of V1 corresponding to increased usage. However, when we implemented an individual-specific approach, we observed that both increased and decreased usage leads to alterations in FC patterns. Results suggested that increased usage leads to idiosyncratic changes in connections whereas decreased usage leads to more stereotyped connection patterns. Further, FC patterns to parts of V1 that process peripheral vision are more stereotyped than patterns of connections to central vision (F(1,10877)=8.5697, p=0.0034), whereas they are equally stereotyped in patients with central vision loss (F(1,1447)=0.34051, p=0.56). Thus, more nuanced changes in connections can be observed when inter-individual variability is taken into account. Overall, our study emphasizes the diversity in patterns of brain plasticity following central vision loss and highlights that FC from V1 maintains the capacity to adapt in adulthood.

Acknowledgements: We would like to thank NIH/NEI Grants to Visscher 1 U01 EY025858-01A1 & 1R01EY031589-01.

Talk 5, 6:15 pm

Geometric changes in monkey V4 and IT neural responses during visual category learning

Jenelle Feather1,2 (), Long Sha2, Gouki Okazawa2,3, Nga Yu Lo1,2, SueYeon Chung1,2, Roozbeh Kiani2; 1Flatiron Institute, 2New York University, 3Institute of Neuroscience, Shanghai

Behavior changes over the course of learning a task. This behavioral change is due to shifts in neural responses that support improved performance. Here, we investigated how the underlying representational geometries in visual areas V4 and IT of the Macaque visual system change during a categorization learning task. Visual stimuli varied in two independent attributes, and monkeys learned to categorize them based on a category boundary in the stimulus space that was defined by a combination of the attributes. Chronic neural population recordings were obtained from V4 and IT over multiple days of training while a monkey learned the task through receiving correct/incorrect feedback. Additionally, we recorded from the same neural populations while a monkey performed a fixation task viewing the same sets of stimuli. In all eight analyzed tasks, the monkey’s performance on the categorization task improved with training. To link this behavioral improvement to the underlying population responses, we investigated how the geometry of neural population activity changed over the course of learning. We treated population responses to all stimuli in each of the two categories as manifold-like representations, and analyzed the geometric properties of these representations using mean-field theoretic manifold capacity analysis. As the monkey learned the task, we observed that the representations in both V4 and IT for the two classes became more separable as measured by an increase in manifold capacity. This increase in capacity was associated with a characteristic geometric change in the neural population response geometry. Our results suggest that both V4 and IT responses actively change during category learning in ways that directly lead to increased separability and improved readouts for downstream neural areas, and point towards future work linking these population-level geometric changes to local changes at the single-neuron level.

Talk 6, 6:30 pm

Differential unconscious control of the medial prefrontal cortex during non-REM and REM sleep to mitigate anterograde and retrograde interferences in visual perceptual learning

Takashi Yamada1 (), Theodore LaBonte-Clark1, Shazain Khan1, Peter Sage1, Kalyan Pooja1, Hana Berhe1, Sophia Hopkins1, Yu-Ang Chen1, Yusuke Nakashima1, Takeo Watanabe1, Yuka Sasaki1; 1Brown University

While it is established that visual perceptual learning (VPL) is associated with changes in visual areas, the roles of the prefrontal cortex (PFC) in regulating top-down signals for VPL remain elusive. This study specifically investigates how the PFC unconsciously controls top-down signals during sleep to reduce anterograde and retrograde interferences between VPL of the two distinct visual tasks trained before and after sleep. Before and after a 90-minute sleep session inside an MRI scanner with polysomnography, two interfering texture discrimination tasks (TDT) were trained. We tested whether training of pre-sleep and post-sleep TDTs interfered with each other. As our prior research (Nature Neuroscience, 2020) demonstrated a significant correlation between the concentration of excitatory-to-inhibitory neurotransmitters (E/I ratio) in the visual cortex during REM sleep and resilience to interference, we measured E/I ratios during non-REM and REM sleep in the medial prefrontal cortex (mPFC) and dorsolateral prefrontal cortex (DLPFC) regarding interference. Subjects who exhibited both NREM and REM sleep demonstrated greater resilience to retrograde interference (from post-sleep TDT to pre-sleep TDT) compared to those who showed NREM sleep alone. In mPFC, the E/I ratio significantly reduced from baseline during REM sleep in correlation with resilience to retrograde interference. However, resilience to anterograde interference (from pre-sleep TDT to post-sleep TDT) was correlated with the E/I ratio during NREM sleep. DLPFC exhibited no significant correlations between E/I ratios and resilience to interference anterogradely or retrogradely, while increases in the E/I ratio in DLPFC during NREM sleep from baseline were significantly correlated with offline performance gains in pre-sleep TDT. These findings indicate the involvement of the mPFC in both anterograde and retrograde interferences but during distinct sleep stages. Despite the prevailing belief that consciousness is predominantly involved in prefrontal controls, our findings suggest that prefrontal control mechanisms operate even during unconscious states such as sleep.

Acknowledgements: NIH (R01EY031705, R01EY019466, R01EY027841), KAKENHI (JP20KK0268).

Talk 7, 6:45 pm

Reversal Learning in the Human Visual Cortex

Caitlin, M Traiser, Richard, T Ward, Hannah, M Engle, Andreas Keil; 1University of Florida

Reversal learning paradigms are commonly used to investigate cognitive and affective processes, including in neuropsychiatric conditions. The present study uses a novel aversive reversal learning paradigm to investigate visuocortical responses to threat stimuli, with prior research focusing on limbic and frontocortical regions. Participants (N = 44; 18-23 years) viewed flickering Gabor patches at different orientations, driving steady-state visual evoked responses (ssVEP) recorded with EEG. An aversive loud noise was used as the unconditioned stimulus, consistently paired with one orientation (the CS+) and never with the other (CS-). After the initial acquisition phase, the contingency between the conditioned and unconditioned stimuli was reversed. Test phases following initial acquisition and reversal examined frequency-tagged ssVEPs evoked by CS+ and CS-, as well as a neutral accompanying Gabor, allowing us to quantify competition effects as a function of learning. Participants were asked to rate each stimulus in terms of valence, arousal, and expectancy before, during, and after learning. Continuous EEG was recorded using a saline EEG system with 129 electrodes/sensors and artifact-free trials were analyzed in the frequency domain, using the Discrete Fourier Transform, after averaging trials by condition. Statistical analyses were conducted using Matlab. We compared the ssVEP amplitude at the tagging frequencies during the critical test phases, across the entire topography. As expected, the ssVEP evoked by the conditioned threat cue (CS+) was enhanced over posterior sites, compared to the CS-, after the initial acquisition phase. Importantly, this effect reversed after 60 trials of reversal learning, and increased in effect size: The new CS+ (the former CS-) prompted selectively heightened ssVEP signals compared to the new CS-. Findings support the notion that experience changes the amplitude of neural mass activity in human visual cortex. They also show that these changes are malleable, adapting and even reversing with environmental contingencies.

Acknowledgements: The research was supported by NIH grant R01MH125615 to Andreas Keil, PhD.

Talk 8, 7:00 pm

Visual perceptual learning completely transfers to a new location when phase, instead of contrast, is varied during training

Beyza Akkoyunlu1,2,3,4 (), Caspar Schwiedrzik1,2,3; 1Perception and Plasticity Group, German Primate Center, Göttingen, Germany, 2Neural Circuits and Cognition Lab, European Neuroscience Institute Göttingen – A Joint Initiative of the University Medical Center Göttingen and the Max-Planck-Society, 3Leibniz ScienceCampus Primate Cognition, Göttingen, Germany, 4Graduate School for Neuroscience, Biophysics and Molecular Biosciences (GGNB), Göttingen

Performance on visual tasks can be improved via training or experience, and this is called visual perceptual learning (VPL). However, this improved performance is limited to the trained task’s specifics, i.e. when the spatial position of the stimulus is changed, the improvement disappears. Yet, recent research shows that variability along task-irrelevant stimulus dimensions can alter this characteristic. We argue that variability determines which neurons undergo plasticity in VPL, and depending on these neurons’ invariance properties, generalization or specificity is achieved. We trained two groups of participants with almost identical tasks, only changing which task-irrelevant dimension varied between trials. In particular, we created variability by randomizing spatial phase in one training group and contrast in the other. After training, we tested for transfer to a new spatial location in both groups. Phase-invariant neurons emerge later in the visual processing hierarchy, compared to contrast-invariant neurons (e.g., complex, and simple cells), and hence phase-invariant neurons have larger spatial receptive fields. Due to this, we hypothesized that varying the phase of the training stimuli over trials would give rise to generalization in space due to neurons which are phase invariant taking a role in the training. On the other hand, as contrast-invariant neurons appear earlier in the hierarchy, we expected the learning to be more specific when participants were trained with varying contrast. We found that the randomizing phase of the training stimulus resulted in complete generalization of the improvement to a new spatial location, contrary to randomizing contrast. Our results show that which neural populations undergo plasticity with VPL is determined by the training task demands, and in turn, this affects generalization and specificity of behavioral improvements.

Acknowledgements: This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (Grant agreement No. 802482, named “VarPL”)