Malcolm A. MacIver - US grants

Animals are thought to have diverged from plants more than 1.5 billion years ago. Basic to this split are different strategies for obtaining the energy needed for life: for a plant, it is "stay in place and absorb," and for an animal, it is "move around and grab." As soon as motion enters the scene, so do two quite different regimes under which motion can occur: the first is the "viscous" regime, in which an animal will stop in its tracks as soon as it ceases generating locomotory forces, and the second is the "dynamic" regime, in which an animal will keep moving even after it ceases generating such forces. For our fluid-bound ancestors, this transition occurred with the dawn of the multicellular animals around 0.6 billion years ago. Control of motion is much more difficult in the dynamic regime, a fact well known in the engineering of robotic systems. This sets the fundamental problem for nervous systems to solve: the transformation of sensory signals into motor signals in a manner that accounts for the animal's dynamic constraints. Dr. MacIver will lead a multidisciplinary group of researchers with expertise in neuroscience, robotics, and fluid dynamics to understand how sensory signals are transformed into motor signals by the brain of weakly electric fish, Apteronotus albifrons, with particular attention paid to how the dynamic constraints of the fish affect this transformation. The researchers hypothesize that neural structures supporting this transformation are simplified by sensory and motor capabilities that are well tuned to the dynamics of the task. The team's research objectives are to 1) reconstruct muscle activations occurring during prey-capture behavior; 2) reconstruct the sensory information about the prey reaching the brain during this behavior; and 3) develop a computational framework for transforming the reconstructed brain input into the estimated muscle activation signals. Experiments on real fish and on a virtual fish with realistic sensing and mechanics will be combined to test several key hypotheses, including the claim that a trajectory to the prey that minimizes the animal's effort will be identical to one that minimizes uncertainty about behaviorally relevant properties of the prey, such as its location. These studies require an ambitious interdisciplinary effort in neurobiology, computational neuroscience, fluid dynamics, and robotics. The research will have broad applicability to understanding the principles of sensorimotor transformations in animals. The group further expects that their work on the fluid dynamics of locomotion will have applications to animal flight and swimming, and the engineering of micro-air and aquatic vehicles. The project will also involve undergraduate students in aspects of the research and will develop a robotic fish installation to inform the public about this type of multidisciplinary research.

CBET-0828749
Patankar

The physical principles underlying the extraordinary mobility of swimming and flying animals have been the subject of years of effort and there is still much that is not understood. This study develops an efficient numerical method for fully resolved simulation of self-propulsion of organisms called Fully Resolved Momentum Redistribution for self Propulsion (FuRMoRP). It will be used to study swimming fish; however, it is sufficiently general to function for small flying animals as well. The motivation to develop such a tool is two-fold: first, to develop a high resolution efficient fluid simulation technology that is transformational by its potential to significantly impact many interdisciplinary areas; second, to gain insight into a number of fundamental problems in aquatic locomotion which will also lead to insights into the design of a novel, highly-maneuverable underwater vehicle being developed through a separate project in the Co-PI's lab. The maneuverability and efficiency of fish is inspiring new styles of propulsion and maneuvering in underwater vehicles for applications such as undersea exploration and environmental monitoring. The development of such vehicles will depend on the resolution of open issues in aquatic locomotion which will be studied here using FuRMoRP applied to three important swimming modes and fish morphologies. The PIs hope to answer specific questions: What is the most efficient deformation kinematics three given fish types? How do they compare with experimentally observed gaits? What are their comparative efficiencies? The education plan involves developing new graduate and undergraduate courses, fluid animations for explanation of biofluid-dynamic principles, a book project, and international outreach. For outreach, the PIs will work with the world renowned Shedd Aquarium in Chicago to help develop a more educational display of the electric eels. The display will provide real-time acoustic and visual cues to the visitors to help them appreciate some of the fluid dynamical science and beauty of the electric eels.

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5).

While the nervous system operates with information, the mechanics of the body and the environment in which it is embedded constitute a world of forces. Work on the mechanics of the body and on the nervous system is rarely undertaken in a joint fashion, in part because of the difficulty of comparing these quantities. A theoretical umbrella under which neural information acquisition can be related to mechanics is currently lacking; there is no science of infomechanics. This project will use the model system of weakly electric fish to push forward an understanding of the linkages between obtaining sensory information and movement mechanics. It has recently been shown that, unlike most forward-biased animals which sense objects ahead better than in other directions, electric fish are able to sense in all directions. Complementing this unique sensory capacity is a motor system which allows them reach locations all around the body quickly. The very high degree of coupling between sensation and movement in these animals makes them ideal for elucidating the principles connecting mechanics to sensory processing. The kinematics and dynamics of this unique motor system will be studied through use of a highly stereotyped refuge-tracking behavior and work on an advanced electric fish robot. The neural basis of sensory-guided movement will be examined through a study of the encoding of locomotor signals in the brain. Finally, a robotic model of a closed-loop sensory tracking behavior present in fish will be used to test hypotheses of how the processing of object features is connected with movement control. This work will result in the training of one postdoctoral associate and several graduate students in interdisciplinary research spanning behavior, neurobiology, robotics, and fluid dynamics.

In the traditional view, the nervous system performs the computational "heavy lifting" in an organism. This view neglects, however, the critical role of biomaterials, passive mechanical physics, and other pre-neuronal or non-neuronal systems. Given that neurons consume forty times more energy per unit mass than structural materials such as bone, it is better, when possible, that biological systems employ relatively inexpensive structural materials rather than relying on more costly neuronal control. In this "bone-brain continuum" view, animal intelligence and behavioral control systems can only be understood using integrative modeling approaches that expose the computational roles of both neural and non-neural substrates and their close coupling in behavioral output. To this end, a group of researchers from Northwestern University, The Johns Hopkins University, and Harvard University propose to create a unique high fidelity neuromechanical model of a vertebrate. The effort is divided between the development of a general purpose computational tool set for neuromechanics research and application of these tools to an ideally suited model system, weakly electric knifefish.

The research will lead to breakthroughs in fundamental problems of how nervous systems work together with biomechanics to generate adaptive behavior. The final goal of the research is to construct an integrated neuromechanical model of a unique biological system - weakly electric knifefish - that places biomechanics and neural control on equal footing. Prior such neuromechanical models have used highly simplified models of mechanics and highly abstracted neuronal control approaches. This research advances the state of the art by incorporating high-fidelity mechanics with neuronal mechanisms motivated by direct neurophysiological evidence. Ultimately, this computational approach will help elucidate how animals distribute computations between brain and bone.

Human telepresence underwater is essential for tasks such as security sweeps in harbors and oil field servicing. Co-robotic solutions are needed, because the risks are great for human divers, while autonomous robots do not deal well with contingencies. A major problem is that vision works poorly in murky environments, such as when mud is kicked up from the bottom. In this National Robotics Initiative (NRI) project the researchers are investigating and developing a replacement for vision -- electrosense -- used by Amazonian fish that navigate and hunt in murky water. These "weakly electric fish" generate an AC electric field that is perturbed by objects nearby. Electroreceptors covering the body of the fish detect the perturbations, which the fish decodes into information about its surroundings. The researchers are developing methods of preprocessing electric images for human understanding, and new computed methods for machine interpretation.

The research creates electrosense hardware and practical testbeds, for navigation and for manipulation underwater. It investigates methods and software to facilitate human interpretation of electric images, as well as machine interpretation. In hardware, the researchers are creating a kilopixel-scale electrosense array as an input sensor for human interpretation of electric images, and development of preprocessing algorithms to make human interpretation workable. The researchers are also using sparser and non-coplanar groups of electroreceptors on a manipulator, for control of pre-grasp and manipulation tasks. For human interpretation, electric image preprocessing includes contour painting and spatial high-pass filtering, as well as temporal filtering. For machine interpretation, methods include specific recognition strategies for simple geometric primitives, and sparse beamforming techniques for more complex environments.

The decision to approach or avoid is a fundamental aspect of animal behavior. How this decision is made by networks of motor neurons that are located in the brainstem and spinal cord, and which trigger muscle cell contraction, is still unclear. This project will investigate the neural and mechanical basis of innate decision-making in vertebrates. Studies will be carried out using zebrafish larvae because they undergo a change in their innate decision-making ability during the first few days after hatching. Immediately after hatching, zebrafish larvae have the ability to generate escape behavior in response to threats. Three days later, they add the ability to not only avoid, but to approach and attack small objects. The goal of this project will be to determine how the neural circuitry supporting the decision to escape or approach visually detected objects is organized during development. Graduate students will be trained in the use of cutting-edge electrophysiological, imaging, computational and behavioral techniques uniquely applied in the zebrafish model system. In addition, an outreach program will be designed and implemented to introduce local high school students to basic neurobiological concepts addressed in this project. The program will involve intuitive and interactive experiments using simple robots with circuits that can be modified to create the approach or avoidance behaviors observed in fish.

The investigators will pursue several hypotheses regarding the development of approach and avoidance behaviors in larval zebrafish. Aim 1 will use high-speed videography and automated body tracking to evaluate the hypothesis that the circuitry for approach is not in place until later on in development. The expectation is that only the older larvae will be able to generate kinematically-distinct responses to attractive visual stimuli. Aim 2 will distinguish between two leading possibilities for how reticular circuitry mediates approach and avoidance, either via the addition of new components or the modification of pre-existing ones. In vivo dye labeling combined with functional calcium imaging approaches will assess changes in the morphologies and responses to visual stimuli of readily identifiable reticular neurons during development. In vivo patch clamp recordings will also be used to confirm outputs to spinal circuitry. Aim 3 will examine the likelihood that newly developed approach circuitry in the spinal cord is either overpowered or shut off by reticulospinal drive during avoidance maneuvers. The predictable read-outs of either scenario will be assessed using electrophysiological recordings of motor output, advanced computational fluids and body modeling, and targeted laser ablations followed by kinematic analysis.

We think of nervous systems as the means by which an animal organizes its world, but a deep time perspective suggests that it is rather the world of an animal that organizes its brain. Prior to the vertebrate invasion of land 385 million years ago, vision, our most powerful long-range sense, took in a largely blurry world at short range while underwater, with little variability in scene as the eyes move. Once on land, vision takes in a high contrast world at long range, with high variability as the eyes move. A possible reason for the greatly increased size and complexity of terrestrial vertebrate brains over those of fish is that this environment provides selective advantage to long sequences of actions toward distant goals, reaching its most complex form in varieties of prospective cognition in certain mammals and birds. A team of Northwestern researchers will conduct research into the computational, behavioral, and neural basis of planning, rooted in an evolutionary and computational sensory ecology perspective and a commitment to ethologically relevant behaviors. Planning is an immensely important capacity to understand the mechanistic basis of, as it participates in a diverse range of behaviors, and its diminishment favors impulsivity and reliance on the habit system. Up to now, laboratory studies of planning have typically relied on reduced environments and simple behaviors which are either appetitive or (more rarely) aversive, without a sentient target, the dynamics and unpredictability of which is likely key to the adequate analysis of prospective cognition. Methods from neuroengineering and data-intensive neuroscience will be brought to bear on the problem of making a more ethologically relevant, yet tightly controlled approach to investigating planning possible. The computational and behavioral work will be used to guide neurobiological interventions in two of the key brain structures that participate in reactive versus reflective decision making and choice: the striatum and hippocampus.

The team will pursue research with an unusually bold intellectual dynamic range well beyond a typical disciplinary approach, from its motivation rooted in evolutionary biology and computational sensory ecology, to the extension of the latest machine learning methods, through to single-cell resolution imaging of live animal behavior in a virtual reality system. The researchers will knit together parallel synergistic efforts in the simulation of planning, a mechatronically reconfigurable behavior arena with a robot predator, and two-photon single cell resolution imaging in a virtual reality system, resulting in an ethologically relevant context significantly more complex than current practice in laboratory settings. There are few areas of neuroscience that have as much potential to impact society as research on the neural basis of planning. Discussions of self-control, marshmallow tests, grit, and challenges we face in making long term plans such as retirement or adapting to changing climate for future generations fill the media. One of the team's research goals is to understand the manner in which the nervous system participates in constraining the temporal and spatial range of prospective cognition,which is clearly quite limited even in humans, toward a neuroscience of sustainability.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

Planning is centrally important for everyday life. Planning is challenging to study, as it involves an internal search through future possibilities for action, in the absence of any sign this is occurring from the outside. The project will investigate two synergistic aspects of real-time planning: its biology and technology. Current artificial intelligence methods require vast amounts of power, a large amount of training, and examination of millions of possible futures to tackle simple problems such as the next move in a turn-based board game. In contrast, mammals require very little power, little to no training, and examination of few possible futures to tackle complex problems such as where to go to hide from a stalking predator. This project will develop a new online planning agent—a robotic “predator”—that will interact with animals trained to evade it inside a complex habitat. The robot will interact with laboratory animals whose brain activity is being recorded while they are challenged—via specially-designed complex habitats—to employ strategic behaviors in avoiding the robot. This will test and advance theory of neural mechanisms underlying the everyday ability to plan in real time in an energy-efficient manner.

The ability to plan actions can produce much larger rewards than reactive, reflexive, or habitual behaviors. Whereas humans exhibit great proficiency in planning and executing daily movements, poor response to long-term threats shows its limits. Research on multi-step planning is in its infancy, constrained in part by behavioral tasks with low ecological validity. Theory has advanced due to rapid progress in artificial intelligence, but most formalizations require so much computing power that real-time planning is impossible. Animals seem likely to form real-time plans in some other way. In prior work, the PIs showed that a selective benefit of visually guided planning may have facilitated the transition onto land 380 million years ago because animals can see targets much farther in air than through water. The benefit of planning in predator-prey engagements is maximized in habitats that afford long sightlines while also providing obstacles that can hide adversaries. In these conditions, such as savanna-like habitats where hominins first emerged, planning its peak advantage. In Aim 1 of the project, this idea is modeled to identify locations of maximal planning payoff (via a network connectivity measure) and used to predict neural computation in animals. This initial algorithm is 10,000 times faster in achieving the same survival rate of simulated prey than a leading competitor in machine learning. This enables creation of a behavioral assay in which live animals are challenged by an adversary with similar planning abilities to their own. With the principle translated into hardware, a bidirectional benefit will emerge for Aim 2. First, neural activity—using Neuropixels probes in freely behaving mice—will be compared to the team’s theory predictions in real-time; they predict that boundary detection cells in the hippocampus and delay interval cells in entorhinal cortex are important for trimming the neurocomputational burden of plans. Second, during recordings, animals will engage with a robot that plans in real time.

This project is funded by Integrative Strategies for Understanding Neural and Cognitive Systems (NCS), a multidisciplinary program jointly supported by the Directorates for Biology (BIO), Computer and Information Science and Engineering (CISE), Education and Human Resources (EHR), Engineering (ENG), and Social, Behavioral, and Economic Sciences (SBE).

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.