Termine / Dates
Sofern nicht anders angegeben, finden alle nachfolgenden Veranstaltungen im Cognium (Hochschulring 18), Raum 2030, um 17 ct statt.
13. Februar 2017
Prof. Dr. József Fiser, Central European University, Budapest
Beyond Hubel & Wiesel: Implications of a probabilistic approach to visual perception, learning and development
Almost 60 years ago the seminal papers of Hubel and Wiesel on the functional architecture of the cat striate cortex opened new vistas in sensory information processing in the brain. These studies established the concepts of receptive fields, visual areas, pipeline feed-forward processing in the visual cortex, segregation of different streams of information within and across visual areas and have fundamentally determined the dominant approaches to sensory information processing in the brain for the following half century. However, after having great initial success, this framework could provide neither convincing explanations to a large number of cortical phenomena nor a satisfactory link between sensory processing and higher-level cognition, initiating a vigorous research for alternative frameworks. I will present one such framework based on the two concepts that a) the brain performs probabilistic computation to generate our perception, and b) that neural responses provide samples from a probability distribution representing the brains best estimate of the present situation. I will highlight the fundamental differences between the classical framework and this probabilistic one, discuss the implications of the probabilistic approach on learning and development in the brain and demonstrate the viability of the framework thought a number of physiological and modeling experiments. Specifically, I will show that the framework creates an essential link between perception and learning, that it implies that spontaneous activity is not noise but the brain’s ongoing momentary estimate about its environment, I will demonstrate how the framework provides explanation of task-dependent correlational activity in V1 without evoking the concept of attention, and how the effect of interference with normal visual development can be captured by this framework. Together these results indicate how the probabilistic framework can offer a comprehensive treatment of perception learning and structural development in the visual system. If time permits, we will also cover nuts and bolts, i.e. the questions of whether sampling-based probabilistic inference really works, and if yes, how.
18. Oktober 2016
Prof. Dr. Michael Herzog, École Polytéchnique Fédérale de Lausanne, Switzerland
A fresh, old look on vision
In classic models of vision, vision proceeds in a feedforward and hierarchical fashion, from low-level analysis (edges and lines) to figural processing (shapes and objects). Low-level processing fully determines high-level processing. For example, the responses of a neuron sensitive to a square are fully determined by the neural activity of neurons coding for the vertical and horizontal edges making up the square. First, using crowding as a paradigm, I will show that shape processing determines low-level visual processing as much as the other way around. For example, we presented a vernier stimulus and asked observers to indicate its offset direction. Performance strongly deteriorated when the vernier was surrounded by a square, in line with most models of vision. Surprisingly, performance improved when more squares were added. All classic models of vision predict just the opposite, i.e., a further deterioration of performance. Second, using visual masking, EEG, and transcranial magnetic stimulation (TMS), I will show that features are unconsciously integrated for a substantial time before consciousness is reached. During unconscious processing, shape processing interacts with low-level processing in a recurrent fashion. This period may last as long as half a second. Third, I will quickly sketch a computational model, which is very little related to the classic models of vision, employing recurrent connections. Computer simulations will show how neural dynamics explain the results of the crowding experiments mentioned above (cooperation with Greg Francis and the Human Brain Project). Even though the model is state of the art with millions of neural connection, it reverberates the tradition of Gestalt theory.
14. Dezember 2015
Prof. Dr. Cornelius Schwarz, Werner Reichardt Center for Integrative Neuroscience, Hertie Institute for Clinical Brain Research, University of Tübingen
The Slip Hypothesis – Tactile Coding and Perception
The haptic sense is a 'scanning' sense, which actively deploys energy to the environment, and indirectly uses the reflected transformation of this energy to build a percept. In this sense whisker and hand movements are better compared to echo-location and electro-sensation, than to hearing and vision. The scanning characteristics prompt special questions about invariance (how to safely identify the same object with different whisker movements) and optimization (how to move to improve perceptual decisions).
To begin to answer these questions, we will have to describe the properties of energy transformation in scanning and to describe the neuronal mechanisms controlling sensory signal flow in movement-dependent ways. I will detail our recent results from rat psychophysics and whisker biomechanics, which speak to an important facet of energy transformation when tactilely scanning: frictional movements or stick-slip movements. We found that frictional movements, and object discrimination derived from them, are systematically dependent on how a whisker is rubbed across a texture. A first set of psychophysical experiments approximated frictional movements with pulsatile whisker deflections. The results from these experiments are in line with the hypothesis that rats analyze vibrotactile signals in an instantaneous way, which seems appropriate to extract the waveforms of rare information carrying events embedded in noise. As frictional movements are an integral element of all moving contacts, they have a great potential to play a unifying role to build a common perceptual hypothesis for whisker- as well as finger-related sensing.
7. Dezember 2015
Prof. Dr. Jannis Hildebrandt, Universität Oldenburg
Two Sides to Every Coin: Trade-offs in Adaptive Coding in the Auditory System
Encoding of sensory stimuli changes according to both environment and current internal requirement of the animal. Often, sensory adaptation is seen as a process of optimization. Muss less appreciated is the fact that changes to stimulus encoding may be beneficial for representation of one aspect of the stimulus, but deteriorate information about other - possibly important - features. I will demonstrate this trade-off in examples from the auditory systems of two distant animal groups: grasshoppers and the mammalian auditory cortex. In grasshoppers, peripheral neural adaption enables invariant sound representation but hinders localization of the stimulus. In auditory cortex, recent evidence suggests that attentional modulation synchronizes activity of nearby cells via activation of specific classes of inhibitory cells. I will show that correlation of activity in the auditory cortex degrades population representation of sound frequency but goes along with more precise encoding of temporal features. This possibly reflects specific demands of the auditory domain, where much of the relevant information resides in the time course of the stimulus.
16. November 2015
Prof. Dr. Martin Greschner, Universität Oldenburg
Structure and Origins of Correlations in the Retina
For many years it has been known that retinal ganglion cells exhibit substantial correlated firing: a tendency to fire nearly synchronously at rates different from those expected by chance. These correlations suggest that network interactions significantly shape the visual signal transmitted from the eye to the brain. However, visual signals are conveyed to the brain by multiple morphologically and functionally distinct types of ganglion cells in parallel, and less is known about how correlated firing influences visual signals across and between these visual pathways. We examined the degree and structure of correlated firing among ganglion cell types and related the different aspects of correlations to several sources: The dominant correlations between nearby ganglion cells were largely consistent with an origin in shared noise from photoreceptors. At scotopic light levels additional slow correlations were present. We examined the correlations introduced by electrical synapses between ganglion cells and studied the long range correlations between ganglion cells and spiking inhibitory interneurons. Finally, the local correlations in spontaneous activity combined with correlations induced by common visual stimulation, and limited the fidelity with which visual signals were encoded by populations of ganglion cells. In summary, while shared photoreceptor noise and stimulus driven activity dominate the overall correlated firing, electrical coupling and interactions with inhibitory interneurons strongly affect cell type specific correlations.
9. November 2015
Dr. Isabelle Bülthoff, Max-Planck Institut für biologische Kybernetik, Tübingen
Race and Identity: What Gives a Face its Race of and What do we Remember About Familiar Faces?
We are very good at classifying familiar and unfamiliar faces according to their race or sex. We also show robust identification capabilities for familiar faces. In this talk, I will present two studies investigating what is important in a face for race classification and what we rely on for person identification. In the first study, I will show that the eyes and the texture (skin) of a face are major determinants of ethnicity for both Asian and Caucasian participants. In the second study, I will show that we remember accurately only the idiosyncratic features of familiar faces, but not their detailed race and sex information. The implications of these results for models of face representation will be presented.
6. November 2015
Cognium, room 1030, 16 ct
Prof. Dr. Heinrich Bülthoff, Max-Planck Institut für biologische Kybernetik, Tübingen
Flying Robots and Flying Cars
26. Oktober 2015
Dr. Jean-Claude Dreher, Research director, CNRS, Neuroeconomics, Reward and decision making team, Centre de Neurosciences Cognitives, France
From Reward Processing to Social Decision Making: Insights from Intracranial Recordings and Model-Based fMRI Studies in Humans
Our laboratory investigates the neural mechanisms underlying decision making, motivation and reward processing in humans, using concepts from cognitive neuroscience, psychology and behavioral economics. We use experimental tools such as model-based functional Magnetic Resonance Imaging, intracranial electrophysiological recordings and pharmacological manipulations to understand the computational processes involved when making a choice. Our goals are to understand the functional organization of the prefrontal cortex in humans, the various functions that the reward dopaminergic system exerts on cognition and motivation and the neural mechanisms underlying dysfunctions of these two systems in patients with neurological or psychiatric illnesses (Parkinson’s disease, patients with focal prefrontal cortex lesions, schizophrenia and pathological gambling). In parallel, we are also studying how individual variations in hormones and genes influence reward processing and decision-making. I will present recent results characterizing how reward types and probability engage specific parts of the orbitofrontal cortex, using intra-cranial recordings (iEEG) in patients with epilepsy and using fMRI in healthy subjects. I will also show how different types of rewards and punishments modulate specific brain systems when learning stimuli-outcome associations. In the social domain, we have recently characterized how the human brain monitors and updates dominance status of others during competitive interactions. Together, our results shed light on the common neurocognitive mechanisms engaged in learning stimuli-outcome associations during non-social situations as well as during social interactions.
29. Juni 2015
Prof. Dr. Lutz Jäncke, University Zürich, Institute of Psychology, Division of Neuropsychology
The Brain of Synesthetes: When Tones Taste Bitter
Synesthesia is a rare but a nevertheless fascinating perceptual phenomenon, which is characterized by an inducing stimulus (and the associated perception) eliciting concurrent perceptions. Common forms of synesthesia are grapheme-color or tone-color synesthesia. In grapheme-color synesthetes the inducer could be a particular letter (e.g., an “A”), which induces a concurrent perception (the color perception). For tone-color synesthetes a particular tone (e.g., a “FIS”) acts as inducer and induces the concurrent color perception. The color perception in both synesthesia variants is the so-called concurrent. Many if not most of the synesthetes experience more than one form of synesthesia, thus they are multi-synesthetes. In addition, most of the synesthetes demonstrate multi-modal synesthesias, in which the inducer perception and the concurrent perception belong to two different sensory modalities e.g., vision and auditory modality. Because of these extraordinary abilities, it comes as no surprise that synesthetes perceive their environment partly differently than ordinary people. For example, music can evoke colored images in tone-color synesthetes while grapheme-color synesthetes will perceive colors when reading books. Meanwhile many variants of synesthesia have been reported. In principle every sense can theoretically be coupled with any other sense. An important point to emphasize is that synesthesia is not the product of imagination or hallucination. It is a very specific perceptual variant, which most likely depends on a particular underlying neural architecture of the involved neural networks.
In my lecture, I will present recent data of my lab showing that the brain of synesthetes is different in terms of the anatomical and neurophysiological underpinnings. Most of the available data support the idea that the brain of synesthetes is anatomically and neurophysiologically highly interconnected providing the basis for this extraordinary ability. In this context I will also discuss the possible implications of this specific ability for these synesthetes and for the understanding of the human brain and for particular psychological functions.
18. Mai 2015
Prof. Dr. Eckart Altenmüller, Institut für Musikphysiologie und Musiker-Medizin HMTM-Hannover
Making Music as a Model for Neuroplasticity: New Findings and Doubts
Sensory-motor skills of musicians have some specific qualities: learning begins at an early age in a playful atmosphere. Routines for stereotyped movements are rehearsed for extended periods of time with gradually increasing degrees of complexity. Via auditory feedback, the motor performance is extremely controllable by both, performer and audience. All movements are strongly linked to emotions, – pleasure or anxiety - , processed by the limbic system. These specific circumstances seem to play an important role for plastic adaptation at several levels of the central nervous system.
There is a dark side to the increasing specialisation and prolonged training of modern musicians, namely loss of control and degradation of skilled hand movements, a disorder referred to as musicians’ cramp or focal dystonia. The first historical record, from 1830, appeared in the diaries of the ambitious pianist and composer Robert Schumann. As was probably the case for Schumann, prolonged practice and pain syndromes due to overuse can precipitate dystonia, which is developed by about 1% of professional musicians and frequently ends their career. Neuroimaging studies point to dysfunctional (or maladaptive) neuroplasticity as its cause.
In this lecture, I will present new data on functional plasticity in musicians casting some doubt on the quality of “everlasting plasticity” of the brain. It seems that early training leads to clearly more efficient networks as compared to late start of training.
We furthermore have new data concerning the etiology of Musician’s dystonia. It seems that there are different categories of motor disturbances either linked to psychological traits, or to basal ganglia dysfunction. I will discuss the role of anxiety, reinvestment and dysfunctional plasticity. Finally, I will present a model accounting for our findings concerning the triggering factors of musician’s dystonia!
11. Mai 2015
Prof. Dr. Barry Lee, Sunny College of Optometry, State University of New York
The Primate Visual Pathway and Color Vision: A Historical View
It is now established that primates are the only mammals to have evolved full trichromatic vision . A first suggestion as to chromatic coding in the visual pathway of primates was that the six layers of the lateral geniculate nucleus (LGN) was devoted to different cone types, one per eye per cone (Le Gros Clark, 1949); recent work on koniocellular laminae has suggested this principle is not so outlandish. But soon the presence of color opponency in the primate visual pathway became established. I shall review the development of ideas about color coding in the primate, ideas still influence views of primate visual processing held today. I shall also pick out ten issues about the primate visual pathway that are unresolved or underappreciated. There is cumulative evidence that the primate visual pathway has been heavily modified to accommodate trichromacy, whatever the phylogenetic origins.
Le Gros Clark, W. (1949). Doc. Ophthalmol. 3, 57-64.
27. April 2015
Prof. Dr. Jochem Rieger, Institute of Psychology, Carl-von-Ossietzky University, Oldenburg
Statistical Learning for Neuroscience and BCI
Establishing tight links between neural activation and subjective experience or behavioral variables is an important goal of cognitive neuroscience. I will argue that statistical learning techniques, well known from Brain Computer Interfacing, can provide a powerful approach to tackle this problem.
In my lab we study human neural function along the perception action cycle using intracranial recordings (ECoG) and fMRI and the work I present is organized along these topics. I will start with work on speech perception using intracranial recordings in humans to demonstrate an isomorphism between brain responses and subjective speech perception, extend the approach to characterize the neural substrate of decision making and to track online subjective visual perception using real time fMRI and conclude with recent work on analyzing the cortical control of the upper limb movements.
20. April 2015
Prof. Dr. Helmut Hildebrandt, Klinikum Bremen-Ost, Zentrum für Neurologie & Universität Oldenburg, Institut für Psychologie
Fatigue bei Multipler Sklerose
Fatigue ist ein häufiges Begleitsymptom der Multiplen Sklerose und stellt eine durchgreifende Minderung der Lebensqualität dar. Bis heute sind die Ursachen von Fatigue nicht vollständig geklärt. Die vorliegenden Daten deuten einerseits auf eine starke psychische Komponente, denn Fatigue tritt auch im Rahmen von Depression auf. Ähnlich wie bei der Fatigue sind auch depressive Symptome bei Patienten mit Multipler Sklerose erhöht. Andererseits gibt es auch Hinweise auf die Verursachung durch erhöhte kompensatorische Anstrengung infolge vorhandener Hirnatrophie. Als dritte Ursache kommt eine Störung der hypothalamischen Funktionen in Frage, die z.B. durch das Entzündungsgeschehen verursacht sein könnte. Dann wäre Fatigue z.B. als Teil des Sickness Behaviour zu erklären.
In dem Vortrag werden verschiedene Forschungsergebnisse unserer Arbeitsgruppe referiert, die sich um die dargestellten Erklärungstheorien drehten. Darauf aufbauend wird ein Modell entwickelt, welches die verschiedenen empirischen Ergebnisse der Fatigueforschung integriert und damit unsere weitere Forschung ausrichten soll.