Distinct rhythmic locomotor patterns can be generated by a simple adaptive neural circuit: Biology,simulation, and VLSI implementation

Rhythmic motor patterns can be induced in leg motor neurons of isolated locust thoracic ganglia by bath application of pilocarpine. We observed that the relative phases of levators and depressors differed in the three thoracic ganglia. Assuming that the central pattern generating circuits underlying these three segmental rhythms are probably very similar, we developed a simple model circuit that can produce any one of the three activity patterns and characteristic phase relationships by modifying a single synaptic weight. We show results of a computer simulation of this circuit using the neuronal simulator NeuraLOG/Spike. We built and tested an analog VLSI circuit implementation of this model circuit that exhibits the same range of behaviors as the computer simulation. This multidisciplinary strategy will be useful to explore the dynamics of central pattern generating networks coupled to physical actuators, and ultimately should allow the design of biologically realistic walking robots.     

Bearbeitergesteuerte Analyse von Verhaltenszeitreihen am Digitalrechner

1. Both the organization of behaviour and communicative interactions can be established by analyzing behavioural time series. By means of an analysis of this kind, conclusions about the control of behaviour and the principles of interindividual communication can be reached. An analysis may be based on simultaneous and successive behavioural events. In these multi-channel time series the temporal arrangement of patterns (“Strukturierung”), the organization and the conditions determining the occurrence of behavioural events can be specified.
2. Data-analysis by digital computer permits rapid processing of a large quantity of material in regard to several respects. “Sets of data” can be stored, thus enabling the user to correct, alter, combine and finally analyze them with respect to various questions. For this purpose, a program system for the digital processing of behavioural time series (title “PROVED”) was developed.
3. The course of behavioural events is composed of temporal patterns which are ordered in a definite hierarchy (Fig. 1). Several quantitative and qualitative characteristics (“Muster-Merkmale”) can be attributed to each pattern (Figs. 2,4). With the PROVED-System, the temporal arrangement of behavioural patterns is reestablished by the computer according to the control orders of the user. These orders must take into consideration the physiological relevance that is to be confirmed by an analysis of time serics (Fig. 3). The orders controlling the temporal arragement as well as the asignment of pattern characteristics to sets of data are available for further evaluation.
4. The PROVED-System includes several procedures involving input, output, storage, administration, correction, assignment of characteristics, temporal arrangement and analysis of data. The examiner can select and combine these procedures at will using statements of a particular control language.
5. During the first step of data analysis, a fixed number of parameters for each pattern is determined, the number of parameters varying with the complexity of the problem. Any measured value which is related to the pattern under investigation can be regarded as a parameter. During the second step, multidimensional frequency distributions of crossclassified parameters are established which can be examined by means of statistical tests (third step) (Figs. 5,6).
6. Determination of the parameters by the examiner decisively influences the further course of the analysis. The control statements for one parameter consist of two parts: 1. quality of the parameter (type of pattern, temporal distance, etc.), 2. position of the parameter in the behavioural time series (pattern under investigation, immediately subsequent pattern, next pattern with the same characteristics, etc.). The latter is determined by a series of shifts of indices controlled by the user (Fig. 7,8).
7. The PROVED-System can be applied to the analysis of various behavioural time series (succession of sounds, recordings of movements, observations of behavioural acts, etc.). Depending upon the aim of the analysis (temporal arrangement, organization, control of behaviour or communication), spontaneous behavioural events, input-output experiments (stimulus-reaction) and social interaction can be evaluated.

     

Pattern recognition software and techniques for biological image analysis

The increasing prevalence of automated image acquisition systems is enabling new types of microscopy experiments that generate large image datasets. However, there is a perceived lack of robust image analysis systems required to process these diverse datasets. Most automated image analysis systems are tailored for specific types of microscopy, contrast methods, probes, and even cell types. This imposes significant constraints on experimental design, limiting their application to the narrow set of imaging methods for which they were designed. One of the approaches to address these limitations is pattern recognition, which was originally developed for remote sensing, and is increasingly being applied to the biology domain. This approach relies on training a computer to recognize patterns in images rather than developing algorithms or tuning parameters for specific image processing tasks. The generality of this approach promises to enable data mining in extensive image repositories, and provide objective and quantitative imaging assays for routine use. Here, we provide a brief overview of the technologies behind pattern recognition and its use in computer vision for biological and biomedical imaging. We list available software tools that can be used by biologists and suggest practical experimental considerations to make the best use of pattern recognition techniques for imaging assays.     

A feedback quenched oscillator produces turing patterning with one diffuser

Efforts to engineer synthetic gene networks that spontaneously produce patterning in multicellular ensembles have focused on Turing's original model and the "activator-inhibitor" models of Meinhardt and Gierer. Systems based on this model are notoriously difficult to engineer. We present the first demonstration that Turing pattern formation can arise in a new family of oscillator-driven gene network topologies, specifically when a second feedback loop is introduced which quenches oscillations and incorporates a diffusible molecule. We provide an analysis of the system that predicts the range of kinetic parameters over which patterning should emerge and demonstrate the system's viability using stochastic simulations of a field of cells using realistic parameters. The primary goal of this paper is to provide a circuit architecture which can be implemented with relative ease by practitioners and which could serve as a model system for pattern generation in synthetic multicellular systems. Given the wide range of oscillatory circuits in natural systems, our system supports the tantalizing possibility that Turing pattern formation in natural multicellular systems can arise from oscillator-driven mechanisms.     

Color pattern analysis of nymphalid butterfly wings: revision of the nymphalid groundplan

To better understand the developmental mechanisms of color pattern variation in butterfly wings, it is important to construct an accurate representation of pattern elements, known as the "nymphalid groundplan". However, some aspects of the current groundplan remain elusive. Here, I examined wing-wide elemental patterns of various nymphalid butterflies and confirmed that wing-wide color patterns are composed of the border, central, and basal symmetry systems. The central and basal symmetry systems can express circular patterns resembling eyespots, indicating that these systems have developmental mechanisms similar to those of the border symmetry system. The wing root band commonly occurs as a distinct symmetry system independent from the basal symmetry system. In addition, the marginal and submarginal bands are likely generated as a single system, referred to as the "marginal band system". Background spaces between two symmetry systems are sometimes light in coloration and can produce white bands, contributing significantly to color pattern diversity. When an element is enlarged with a pale central area, a visually similar (yet developmentally distinct) white band is produced. Based on the symmetric relationships of elements, I propose that both the central and border symmetry systems are comprised of "core elements" (the discal spot and the border ocelli, respectively) and a pair of "paracore elements" (the distal and proximal bands and the parafocal elements, respectively). Both core and paracore elements can be doubled, or outlined. Developmentally, this system configuration is consistent with the induction model, but not with the concentration gradient model for positional information.     

Identification of nonlinear systems using random impulse train inputs

Nonlinear systems that require discrete inputs can be characterized by using random impulse train (Poisson process) inputs. The method is analagous to the Wiener method for continuous input systems, where Gaussian white-noise is the input. In place of the Wiener functional expansion for the output of a continuous input system, a new series for discrete input systems is created by making certain restrictions on the integrals in a Volterra series. The kernels in the new series differ from the Wiener kernels, but also serve to identify a system and are simpler to compute. For systems whose impulse responses vary in amplitude but maintain a similar shape, one argument may be held fixed in each kernel. This simplifies the identification problem. As a test of the theory presented, the output of a hypothetical second order nonlinear system in response to a random impulse train stimulus was computer simulated. Kernels calculated from the simulated data agreed with theoretical predictions. The Poisson impulse train method is applicable to any system whose input can be delivered in discrete pulses. It is particularly suited to neuronal synaptic systems when the pattern of input nerve impulses can be made random.     

Cat and monkey cortical columnar patterns modeled by bandpass-filtered 2D white noise

A simple algorithm based on bandpass-filtering of white noise images provides good quality computer reconstruction of the cat and monkey ocular dominance and orientation column patterns. A small number of parameters control the frequency, orientation, branchedness, and regularity of the column patterns. An oriented (anisotropic) bandpass filter followed by a threshold operation models the macaque ocular dominance column pattern and cat orientation column system. An unoriented (isotropic) bandpass filter models the cat ocular dominance column pattern and the macaque orientation column system. The resemblance of computer graphic simulations produced by this algorithm and histological pattern data, is strong. Since this algorithm is very fast, we have been able to extensively explore its parameter space in order to determine filter parameters which closely match the structure of the various cortical systems. In particular, we have applied spectral analysis to our recent computer reconstruction of the macaque ocular dominance column system, and the model produced by the present algorithm is in close agreement with this detailed data analysis.This work was supported by AFOSR 88-0275, the Nathan Kline Psychiatric Institute and the System Development Foundation     

Expert systems in treating substance abuse.

Computer programs can assist humans in solving complex problems that cannot be solved by traditional computational techniques using mathematic formulas. These programs, or "expert systems," are commonly used in finance, engineering, and computer design. Although not routinely used in medicine at present, medical expert systems have been developed to assist physicians in solving many kinds of medical problems that traditionally require consultation from a physician specialist. No expert systems are available specifically for drug abuse treatment, but at least one is under development. Where access to a physician specialist in substance abuse is not available for consultation, this expert system will extend specialized substance abuse treatment expertise to nonspecialists. Medical expert systems are a developing technologic tool that can assist physicians in practicing better medicine.     

Branching tree model with fractal vascular resistance explains fractal perfusion heterogeneity

Perfusion heterogeneities in organs such as the heart obey a power law as a function of scale, a behavior termed "fractal." An explanation of why vascular systems produce such a specific perfusion pattern is still lacking. An intuitive branching tree model is presented that reveals how this behavior can be generated as a consequence of scale-independent branching asymmetry and fractal vessel resistance. Comparison of computer simulations to experimental data from the sheep heart shows that the values of the two free model parameters are realistic. Branching asymmetry within the model is defined by the relative tissue volume being fed by each branch. Vessel ordering for fractal analysis of morphology based on fed or drained tissue volumes is preferable to the commonly used Strahler system, which is shown to depend on branching asymmetry. Recently, noninvasive imaging techniques such as PET and MRI have been used to measure perfusion heterogeneity. The model allows a physiological interpretation of the measured fractal parameters, which could in turn be used to characterize vascular morphology and function.     

Taking the example of computer systems engineering for the analysis of biological cell systems

In this paper, we discuss the potential for the use of engineering methods that were originally developed for the design of embedded computer systems, to analyse biological cell systems. For embedded systems as well as for biological cell systems, design is a feature that defines their identity. The assembly of different components in designs of both systems can vary widely. In contrast to the biology domain, the computer engineering domain has the opportunity to quickly evaluate design options and consequences of its systems by methods for computer aided design and in particular design space exploration. We argue that there are enough concrete similarities between the two systems to assume that the engineering methodology from the computer systems domain, and in particular that related to embedded systems, can be applied to the domain of cellular systems. This will help to understand the myriad of different design options cellular systems have. First we compare computer systems with cellular systems. Then, we discuss exactly what features of engineering methods could aid researchers with the analysis of cellular systems, and what benefits could be gained.     

Taking the example of computer systems engineering for the analysis of biological cell systems

In this paper, we discuss the potential for the use of engineering methods that were originally developed for the design of embedded computer systems, to analyse biological cell systems. For embedded systems as well as for biological cell systems, design is a feature that defines their identity. The assembly of different components in designs of both systems can vary widely. In contrast to the biology domain, the computer engineering domain has the opportunity to quickly evaluate design options and consequences of its systems by methods for computer aided design and in particular design space exploration. We argue that there are enough concrete similarities between the two systems to assume that the engineering methodology from the computer systems domain, and in particular that related to embedded systems, can be applied to the domain of cellular systems. This will help to understand the myriad of different design options cellular systems have. First we compare computer systems with cellular systems. Then, we discuss exactly what features of engineering methods could aid researchers with the analysis of cellular systems, and what benefits could be gained.     

ABPL

Computer analysis of biological systems, using approaches such as metabolic control analysis is common. A typical example is a language like Herbert Sauro's SCAMP (Sauro & Fell, 1991), which allows simulations of enzyme systems, and calculation of control coefficients and elasticities. However such systems are motivated by the underlying biochemical theory and often have limitations as programming languages which mean that they can only be applied to particular classes of problems. ABPL (a biochemical programming language) extends these ideas by adding all the facilities of a fully-fledged programming language, together with some of the capabilities of a modern computer algebra system. Syntactically it derives from the programming language LISP, while the underlying functionality is that of iMAP, the successor to SCAMP. This provides us with a computer system capable of performing most of the tasks undertaken by existing packages, but more importantly, a system which can be easily extended into new areas. Key features of the work are:

- Ability to use the language both interactively and as a batch programming language

- Ability to work both symbolically and numerically

- Ability to handle matrices and vectors

- Ability to define and manipulate reaction schemes

- Common techniques are built in to the language

- Ability to add new operations to the language

The implementation is in ANSI standard C for portability.     

Enabling High Grayscale Resolution Displays and Accurate Response Time Measurements on Conventional Computers

Display systems based on conventional computer graphics cards are capable of generating images with 8-bit gray level resolution. However, most experiments in vision research require displays with more than 12 bits of luminance resolution. Several solutions are available. Bit++ ¹ and DataPixx ² use the Digital Visual Interface (DVI) output from graphics cards and high resolution (14 or 16-bit) digital-to-analog converters to drive analog display devices. The VideoSwitcher ³ described here combines analog video signals from the red and blue channels of graphics cards with different weights using a passive resister network ⁴ and an active circuit to deliver identical video signals to the three channels of color monitors. The method provides an inexpensive way to enable high-resolution monochromatic displays using conventional graphics cards and analog monitors. It can also provide trigger signals that can be used to mark stimulus onsets, making it easy to synchronize visual displays with physiological recordings or response time measurements.Although computer keyboards and mice are frequently used in measuring response times (RT), the accuracy of these measurements is quite low. The RTbox is a specialized hardware and software solution for accurate RT measurements. Connected to the host computer through a USB connection, the driver of the RTbox is compatible with all conventional operating systems. It uses a microprocessor and high-resolution clock to record the identities and timing of button events, which are buffered until the host computer retrieves them. The recorded button events are not affected by potential timing uncertainties or biases associated with data transmission and processing in the host computer. The asynchronous storage greatly simplifies the design of user programs. Several methods are available to synchronize the clocks of the RTbox and the host computer. The RTbox can also receive external triggers and be used to measure RT with respect to external events. Both VideoSwitcher and RTbox are available for users to purchase. The relevant information and many demonstration programs can be found at http://lobes.usc.edu/.     

Use of a neural net computer system for analysis of flow cytometric data of phytoplankton populations

Flow cytometry has been used over the past 5 years to begin detailed exploration of the distribution and abundance of picoplankton in the oceans. Light scattering and fluorescence measurements on individual plankton cells in seawater samples allow construction of population signatures from size and pigment characteristics. The use of "list mode" data has made these studies possible, but on-shore analysis of copious data does not permit on-site reexamination of important or unexpected observations, and overall effort is greatly handicapped by data analysis time. Here we describe the application of neural net computer technology to the analysis of flow cytometry data. Although the data used in this study are from oceanographic research, the results are general and should be directly applicable to flow cytometry data of any sort. Neural net computers are ideally suited to perform the pattern recognition required for the quantitative analysis of flow cytometry data. Rather than being programmed to perform analysis, the neural net computer is "taught" how to analyze the cell populations by presenting examples of inputs and correct results. Once the system is "trained," similar data sets can be analyzed rapidly and objectively, minimizing the need for laborious user interaction. The neural network described here offers the advantages of 1) adaptability to changing conditions and 2) potential real-time analysis. High accuracy and processing speed near that required for real-time classification have been achieved in a software simulation of the neural network on a Macintosh SE personal computer.     

Coarse-grained empirical potential structure refinement: Application to a reverse aqueous micelle

Conventional atomistic computer simulations, involving perhaps up to 10⁶ atoms, can achieve length-scales on the order of a few 10s of nm. Yet many heterogeneous systems, such as colloids, nano-structured materials, or biological systems, can involve correlations over distances up 100s of nm, perhaps even 1 μm in some instances. For such systems it is necessary to invoke coarse-graining, where single atoms are replaced by agglomerations of atoms, usually represented as spheres, in order for the simulation to be performed within a practical computer memory and time scale. Small angle scattering and reflectivity measurements, both X-ray and neutron, are routinely used to investigate structure in these systems, and traditionally the data have been interpreted in terms of discrete objects, such as spheres, sheets, and cylinders, and combinations thereof. Here we combine the coarse-grained computer simulation approach with neutron small angle scattering to refine the structure of a heterogeneous system, in the present case a reverse aqueous micelle of sodium-dioctyl sulfosuccinate (AOT) and iso-octane. The method closely follows empirical potential structure refinement and involves deriving an empirical interaction potential from the scattering data. As in traditional coarse-grained methods, individual atoms are replaced by spherical density profiles, which, unlike real atoms, can inter-penetrate to a significant extent. The method works over an arbitrary range of length-scales, but is limited to around 2 orders of magnitude in distance above a specified dimension. The smallest value for this dimension is of order 1 nm, but the largest dimension is arbitrary. This article is part of a Special Issue entitled "Recent Advances in Bionanomaterials" Guest Editor: Dr. Marie-Louise Saboungi and Dr. Samuel D. Bader.     

A dynamic simulation model of tissue growth and cell patterning

The distributions of cells in tissues of experimental chimaeras and mosaics can serve as tests of mechanisms and rules by which single cells organize themselves into complex, multicellular structures during embryogenesis. We have devised a dynamic, computer simulation model of tissue growth and cell patterning which is directly applicable to the analysis of chimaeras and mosaics. In the model, schematized cells possess a small behavioral repertoire and simple rules for the carrying out of these behaviors. Populations of such cells evolve tissue patterns in real-time that are very similar to those seen in experimental animals. In particular, we have modeled the major pattern features seen in amphibian and mammalian eye chimaeras and mosaics. We have demonstrated that cell mixing can be a passive concomitant of interstitial cell division, a result which alleviates the need to postulate active cell mixing in such mammalian systems. We expect this approach to be a valuable addition to methods of pattern analysis in development.     

Mapping the variation in spider body colouration from an insect perspective

Colour variation is frequently observed in spiders. Such variation can impact fitness by affecting the way spiders are perceived by relevant observers such as prey (i.e. by resembling flower signals as visual lures) and predators (i.e. by disrupting search image formation). Verrucosa arenata is an orb-weaving spider that presents colour variation in a conspicuous triangular pattern on the dorsal part of the abdomen. This pattern has predominantly white or yellow colouration, but also reflects light in the UV part of the spectrum. We quantified colour variation in V. arenata from images obtained using a full spectrum digital camera. We obtained cone catch quanta and calculated chromatic and achromatic contrasts for the visual systems of Drosophila melanogaster and Apis mellifera. Cluster analyses of the colours of the triangular patch resulted in the formation of six clusters and three clusters in the colour space of D. melanogaster and A. mellifera, respectively. Significant differences were found between morphs for both visual systems in contrasts between the colour pattern and two backgrounds against which it would be viewed. Yellow spiders showed higher chromatic contrast than white spiders, while white spiders showed higher achromatic contrast. Therefore, there are perceptual differences between V. arenata colour morphs in the visual systems of potential relevant observers which could pose an important selective pressure on this trait. A variation in the contribution of colour channels to the colour pattern observed in colour maps constructed from reflectance values of individual pixels could influence the way the pattern is perceived, and its resemblance to attractive flower signals.     

Eco-hydrological effects of landscape pattern change

Scientists and environmental managers alike are increasingly concerned about landscape pattern change and its effect on hydrological and ecological processes. In this paper, research progress is reviewed and key issues of eco-hydrological effects of landscape pattern change are discussed. There are different eco-hydrological effects with landscape pattern change, and most attention is paid to runoff, water quality, and soil loss. Landscape shape and spatial distribution can change precipitation-runoff processes and lead to the change in runoff yield. Water quality is closely connected with the composition and spatial pattern of source and sink landscapes. Soil erosion systems are usually modified by land use structure and landscape pattern, and soil loss will be either reduced or increased with land use change. In addition, the change of landscape pattern also has potential impacts on climate and soil quality. In future studies, more attention should be paid to comprehensive multi-scale and integrated research of landscape pattern and eco-hydrological processes.     

Receptor-agonist interactions in service-theoretic perspective, effects of molecular timing on the shape of dose-response curves

Service-theoretic concepts and methods, widely used in other fields (e.g., telecommunication and operations research), are useful also in a biochemical setting because the treatment of biocatalysts (enzymes, receptors) as servers and their ligands as customers, based on the established formal methods of service or queuing theory, may lead to insights and results unobtainable by conventional, mass-action-law-based theories. In this article, we apply the service-theoretic approach to receptor-agonist systems and show how by changing the stochastic time pattern of "operationally relevant" point events (e.g., instants of agonist arrival, instants of post-climax agonist departure) a great variety of dose-response curves may be generated, even in very simple reaction schemes, which, according to mass action kinetics, invariably lead to hyperbolic r(A) curves (r and A stand for response and agonist concentration, respectively). The molecular timing inherent to a hyperbolic response system is not optimal: for instance, at the agonist concentration A(50), half of the agonist molecules are rejected ("lost") because of unfortunate timing of the arrival events. The fraction of lost arrivers can be diminished considerably if the arrivals are better timed: "sub-Poisson" arrivals improve the timing and, thus, convert hyperbolic r(A) curves into "lifted" nonhyperbolic ones. Conversely, "super-Poisson" arrivals make the non-optimal timing in hyperbolic response systems even worse and, thus, convert hyperbolic r(A) curves into "depressed" nonhyperbolic ones. Furthermore, under special timing conditions, nonhyperbolic r(A) curves can be generated, which are partly lifted, partly depressed relative to the reference hyperbola, and which resemble in shape well-known nonhyperbolic forms of enzyme and receptor kinetics (negatively cooperative, positively cooperative, and sigmoidal kinetics). In addition unusual (undulatory and sawtooth-like) r(A) curves can be generated solely by changing the temporal pattern of arrival and service completion instants. Virtually any shape of dose-response curves may be obtained by allowing for probability distributions whose characteristic shape varies with their mean; we call such distributions "variomorphic" and apply them to the arrival process of agonist molecules.     

Analysis of pattern recognition by man using detection experiments

This paper addresses the problem of analyzing biological pattern recognition systems. As no complete analysis is possible due to limited observability, the theoretical part of the paper examines some principles of construction for recognition systems. The relations between measurable and characteristic variables of these systems are described. The results of the study are:

1. Human recognition systems can always be described by a model consisting of an analyzer (F _A) and a linear classifier.
2. The linearity of the classifier places no limits on the universal validity of the model. The principle of organization of such a system may be put into effect in many different ways.
3. The analyzer function F _A determines the transformation of external patterns into their internal representations. For the experiments described in this paper, F _A can be approximated by a filtering operation and a transformation of features (contour line filter).
4. Narrow band filtering (comb filter) in the space frequency domain is inadequate for pattern recognition because noise of different bandwidths and mean frequencies affects sinusoidal gratings differently. This excludes the use of a Fourier analyzer.
5. The relations between the measurable variables, which are the probabilities of detection (P _D curves), and the characteristic variables of the recognition system are established analytically.
6. The probability of detection not only depends on signal energy but also on signal structure. This would not be the case in a simple matched filter system.
7. The differing probabilities of error in multiple detection experiments show that the interference is pattern specific and the bandwidth (steepness of the P _D curves) is different for the different sets of patterns.
8. The distance between the reference vectors in feature space can be determined from the internal representation of the patterns defined by the model. Through multiple detection experiments it is possible to determine not only the relative distances between the patterns but also their absolute position in feature space.