Wim van Drongelen - Academia.edu (original) (raw)

Papers by Wim van Drongelen

Abstract This chapter provides an overview of different types of models for studying the activity... more Abstract This chapter provides an overview of different types of models for studying the activity of nerve cells. It summarizes the neuronal models based on the Hodgkin and Huxley (HH) formalism first described in the 1950s. The HH formalism is a deterministic nonlinear four-dimensional model of the membrane potential dynamics and it can easily be simulated with current computational resources. However, in order to make this approach tractable for mathematical analysis, it must be further simplified. One simplification is to linearize it about a fixed point, e.g., using the resting potential to model subthreshold neuronal behavior. Another common simplification is to reduce the four dimensions into one or two dimensions. In the context of the simplification of the HH model, we employ equivalent electronic circuits. The simplified approach is also related to the integrate-and-fire (IF) modeling approach, first proposed by Lapicque more than a century ago, and neuronal resonance behavior. This chapter includes several MATLAB® routines to simulate the different cellular models and includes instructions on how to build an electronic version of the IF model.

In this chapter we continue the analysis of the filters presented in Chapters 15 and 16 Chapter 1... more In this chapter we continue the analysis of the filters presented in Chapters 15 and 16 Chapter 15 Chapter 16 . Here we consider the filter in the broader context of a linear time-invariant system that was introduced in Chapter 13 . We relate the filter's properties in the time domain and frequency domain. We quantify the relationship between the filter's time constant and the so-called −3-dB point that indicates the transition between pass band and stop band in the filter's frequency response. The filter's frequency characteristic is depicted in two types of plots: the Bode plot and the Nyquist plot. MATLAB ® scripts present examples of the analysis of the filter characteristic and the approach is employed to describe the characteristic measured in Chapter 15 . In addition to the use of sinusoidal or step inputs to perturb the filter, we present and demonstrate the approach where noise is employed to find the filter's frequency response.

Abstract In this chapter we review ordinary differential equations (ODEs) as a tool to model dyna... more Abstract In this chapter we review ordinary differential equations (ODEs) as a tool to model dynamics. We present examples of how to formulate them based on the dynamical system that needs to be modeled, and demonstrate the mathematical techniques one can employ to solve the equation analytically. We show how to solve linear differential equations with and without a forcing term, the so-called inhomogeneous and homogeneous ODEs, respectively. To illustrate the analysis of these equations, ODEs with first-order derivatives (e.g., d c / d t ) and second-order derivatives (e.g., d 2 c / d t 2 ) are used in the examples. Next, we show how higher-order ODEs can be represented as a set of first-order ones, and how this leads to a formalism in matrix/vector notation that can be efficiently analyzed using techniques from linear algebra. To complete the overview of the available tools for solving ODEs, the final part of this chapter briefly refers to application of Laplace and Fourier transforms (see also Chapter 12 ) to solve them.

We detect seizures in newborn infants using a novel method derived from triple correlation, which... more We detect seizures in newborn infants using a novel method derived from triple correlation, which integrates spatial and temporal structure in neonatal electroencephalograms (EEGs). Triple correlation natively encompasses analogues to a variety of lower-order approaches (auto-correlation, cross-correlation) in addition to introducing higher-order signals, so we hypothesized that our approach would both effectively detect and differentiate notoriously difficult-to-detect and heterogeneous neonatal seizures. Indeed, our method in its simplest form performs comparably well to a current standard of care, amplitude-integrated EEG (aEEG), and by some measures outperforms aEEG, suggesting at a minimum that a combination of triple correlation and aEEG could produce a more effective first-line bedside detector. Moreover, we find that the triple correlation seizure-signal varies between patients, with 1) differences in dominance of either within or between channel correlations and 2) differin...

Signal Processing for Neuroscientists, 2018

This chapter discusses the applications of the Fourier transform in spectral analysis and medical... more This chapter discusses the applications of the Fourier transform in spectral analysis and medical imaging. The raw complex-valued output of a fast Fourier transform or discrete Fourier transform is difficult to interpret directly. The most common approach is to present the power spectrum of a given signal. A related approach to displaying the results of spectral analysis is to show amplitude or phase spectra. Spectral analysis is often used in electroencephalography (EEG) analysis to evaluate the classical EEG frequency bands. In physiological signals, interpretation of spectra requires caution because these time series are rarely stationary and usually contain nonperiodic and nonsinusoidal components. The chapter shows examples of the Radon transform, the Fourier slice theorem, and filtered backprojection as each applies to CT image reconstruction. The use of the 2-D Fourier transform and k space are described in the context of magnetic resonance imaging (MRI) application. These techniques require reconstruction of a density function representing the internal structure of an object from sensor readings taken from outside that object.

Signal Processing for Neuroscientists, 2018

Abstract This chapter provides an overview of different types of models for studying the activity... more Abstract This chapter provides an overview of different types of models for studying the activity of nerve cells. It summarizes the neuronal models based on the Hodgkin and Huxley (HH) formalism first described in the 1950s. The HH formalism is a deterministic nonlinear four-dimensional model of the membrane potential dynamics and it can easily be simulated with current computational resources. However, in order to make this approach tractable for mathematical analysis, it must be further simplified. One simplification is to linearize it about a fixed point, e.g., using the resting potential to model subthreshold neuronal behavior. Another common simplification is to reduce the four dimensions into one or two dimensions. In the context of the simplification of the HH model, we employ equivalent electronic circuits. The simplified approach is also related to the integrate-and-fire (IF) modeling approach, first proposed by Lapicque more than a century ago, and neuronal resonance behavior. This chapter includes several MATLAB® routines to simulate the different cellular models and includes instructions on how to build an electronic version of the IF model.

Non UBCUnreviewedAuthor affiliation: The University of ChicagoFacult

Clinical Neurophysiology, 2018

Signal Processing for Neuroscientists, 2010

Gaussian white noise (GWN) allows one to create a series with orthogonal terms that can be estima... more Gaussian white noise (GWN) allows one to create a series with orthogonal terms that can be estimated sequentially with the Lee–Schetzen cross-correlation method. This approach can be adapted when the system's natural input consists of impulse trains such as a spike train. Identifying a system with an impulse train as input is discussed in this chapter. This approach was described by Krausz (1975). Poisson–Wiener operators are orthogonal to all lower-order Volterra operators, which is analogous to the development of the Wiener series with a GWN input. For the generation of a series of random numbers following a Gaussian or a uniform distribution, MATLAB commands randn and rand are used, respectively. A standard MATLAB command for generating a series of intervals according to a Poisson process does not exist. Therefore a Monte Carlo technique is applied to create such an impulse train according to a Poisson process.

Signal Processing for Neuroscientists, 2010

This first chapter describes two of the advanced techniques in signal processing: Lomb's algo... more This first chapter describes two of the advanced techniques in signal processing: Lomb's algorithm and the Hilbert transform. The astrophysicist Lomb developed an algorithm for spectral analysis to deal with signals consisting of unevenly sampled data. The idea of Lomb's algorithm is similar to the development of the Fourier series, namely, to represent a signal by a sum of sinusoidal waves. The Hilbert transform allows one to compute the instantaneous phase and amplitude of a signal. The fact that one can determine these two metrics in an instantaneous fashion is unique because usually this type of parameter can only be associated with an interval of the signal. For example, in spectral analysis the spectrum is computed for an epoch and the spectral resolution is determined by epoch length. Being able to determine parameters such as the phase instantaneously is especially useful if one wants to determine relationships between multiple signals generated within a neuronal network.

Gustatory sensitivity of larvae belonging to nine different Yponomeuta species was studied. A con... more Gustatory sensitivity of larvae belonging to nine different Yponomeuta species was studied. A conspicuous behavioural difference between these closely related species is represented in their host plant preferences. Electrophysiological and morphological examination of the anteriorly-located sensilla revealed that a restricted number of contact chemoreceptors are present. A tip recording technique was used to record gustatory responses of neurones in the lateral and medial styloconic sensilla. The following chemicals were applied as stimuli: sucrose, sedoheptulose, sorbitol, dulcitol, coumarin, phloridzin, salicin, prunasin, rutin, (+)-catechin, the trisodium salt of isocitric acid, ZnCl 2 , NaCl, and KCl. Most of these stimuli were selected on the basis of chemical composition of the array of host plants of the larvae studied (Table 1). Different spike amplitudes displayed in the recordings indicated that most constituents tested are perceived by single cells. Ionized substances sometimes excited more than one cell. In several recordings, a delayed response appeared. This phenomenon seems related to variation in diameter of the distal pore by which inward diffusion of the stimulus takes place. Analysis of the time course of the recordings indicated that the effect of diffusion of stimulus molecules on the initial neural response may be considerable for the non-ionized substances tested. The functional diversity displayed in the interspecific sensitivity patterns can be partly explained by composition of the host plants on which the larvae feed (Table 2). Compounds of one type act as phagostimulants, and some of these compounds are sufficiently host-specific to act as host-recognition factors. A second type of chemical signal inhibits "wrong" food intake; these substances are classified as deterrents. A few gustatory responses appeared non-adaptive, this type of sensitivity is discussed from an evolutionary point of view. 7the maxilla or the maxillary palpi leads to acceptance of normally rejected plants (Torii and Morii 1948; Dethier 1953; Waldbauer and Fraenkel 1961). Assuming a relationship between speciation and host plant choice, the latter studies indicate the significance of the gustatory organs from an evolutionary point of view. Various theories on sympatric and allopatric speciation of phytophagous insects have been described by Dethier (1952), Bush (1974), Wiebes (1976) and Labeyrie (1978). In recent comparative studies on neural and behaviouiral gustatory responses of larvae belonging to the genus Yponomeuta it is suggested that there exists a persistence of sensory sensitivities to constituents of the ancestral host plant (Gerrits-Heybroek et al., 1978; van Drongelen 1978). Because many species of the genus Yponomeuta seem closely related, but show different food regimes (Friese 1960; Herrebout et al., 1976), they are good candidates for a comparative study on insect-host plant relationship. This paper aims to describe gustatory activity during application of plant constituents, as recorded in the lateral and medial styloconic sensilla of nine Yponomeuta species. Parameters of the neural responses are considered in connection with primary events of the peripheral perception mechanism. To evaluate aspects of the relationship between chemosensitivity and occurrence of plant constituents in the host, an attempt is made to state expectations concerning the sensitivity patterns. The distribution of various compounds over the host plants and assumptions on neural mechanisms represent the input of these expectations. Deviations from expected sensitivity are discussed from an evolutionary point of view. A short description of a part of this work appeared previously (van Drongelen 1978) and preliminary data on the sugar receptors have been described by Schoonhoven et al. (1977). Methods Animals. Most larvae of the Yponomeuta species (Lepidoptera, Yponomeutidae) were collected in the field. Larvae of Y. vigintipunctatus and Y. irrorellus were reared in the laboratory on their host plant. The animals were stored in a refrigerator at 6°C for a maximal duration of one month. A list of the Yponomeuta species studied and their associated hosts is shown in the first two rows of Table 1. It can be seen that Y. padellus is oligophagous and the remaining species are monophagous. In this paper, the host-race of Y. padellus feeding on Crataegus species will be referred to as host-race 1 and the one feeding on Prunus spinosa as host-race 2. It was found practical to use the host plant on which the larvae were collected as the main criterion for species identification. On morphological grounds larvae of Y. vigintipunctatus and of Y. plumbellus can be clearly distinguished from those of the other species studied; the first two species belong to a different taxonomie group (group B) than the remaining ones (group A) (Gerrits-Heybroek et al-., 1978). Larvae of Y. cagnagellus were distinguished from those of Y. irrorellus by morphological identification of adult individuals of the latter species, which were reared in a laboratory culture. Morphology. The mouth parts of the insects to be investigated were fixed in Bouin. The fixed preparations were placed in freon liquid, subsequently in liquid nitrogen and were freeze dried. Dried preparations were dissected and coated with gold. The sensory organs were examined in a scanning electron microscope (Jeol, JSM-Ug) in the Service Institute for Technical Physics in Agriculture in Wageningen. Ent. exp. & appl., (in press) CHAPTER 3 Behavioural responses of two small ermine moth species (Lepidoptera: Yponomeutidae) to plant constituents

Chemistry and Physics of Lipids - CHEM PHYS LIPIDS, 2002

Seizure cancellation in a network model with Hodgkin-Huxley type elements was evaluated by applyi... more Seizure cancellation in a network model with Hodgkin-Huxley type elements was evaluated by applying different stimulation paradigms. The simulated activities of single neurons and a computed field potential during different states of the network were the basis of this evaluation. Three types of electrostimulation were investigated: hyperpolarization of the neurons, depolarization, and an electric stimulus that is proportional to the inverted field potential. We demonstrated that a seizure, represented by synchronized network activity, and consequently a large amplitude field potential, could be effectively stopped by the third type of stimulus. The depolarization and hyperpolarization paradigms failed to desynchronize the activities.

Solving differential equations analytically is not always the easiest strategy, or even possible.... more Solving differential equations analytically is not always the easiest strategy, or even possible. In these cases one may use a numerical solution to characterize the dynamics governed by the ordinary differential equation (ODE). A diagram of the flow plotted versus time can create insight into the dynamical behavior governed by the ODE. In addition, the so-called phase space representation, where the dynamics is represented by a vector at each state of the system, is a helpful tool to qualitatively assess system dynamics. Here, we examine both numerical solutions and graphical representations of the dynamics using two growth models: one model with unrestricted growth (a linear ODE), and one where the growth is restricted by limited resources (the logistic equation, a nonlinear ODE). We use different techniques and MATLAB ® examples to demonstrate the numerical approach: Euler's method, the improved Euler's method, and the fourth-order Runge–Kutta algorithm. In the final part we briefly discuss methods for obtaining solutions for partial differential equations.

The Journal of Physiology, Apr 1, 1978

1. Receptor cell activity in the frog's eminentia olfactoria was recorded using m... more 1. Receptor cell activity in the frog's eminentia olfactoria was recorded using metal-filled micro-electrodes. 2. Several units discharged spontaneously with a mean frequency lower than 0.2 spikes per sec, or were silent in periods of up to 5 min. The other units displayed spontaneous activities between 0.2 and 1.05 spikes per sec; their activity could be modelled with a Poisson process. 3. Near-threshold responses to odour stimulation were investigated, considering several stimulations within a small concentration range. Low concentration stimulations were sometimes followed by a response, sometimes not. The concept of response probability is introduced to describe this incertitude. 4. The distribution of the number of spikes in several odour trials at low concentrations showed a reasonable agreement with two types of Poisson distribution. 5. The findings are discussed in connexion with receptor cell sensitivity and the excitation of second order neurones in the bulb.

Journal of Theoretical Biology, Mar 7, 1978

ABSTRACT

Abstract This chapter provides an overview of different types of models for studying the activity... more Abstract This chapter provides an overview of different types of models for studying the activity of nerve cells. It summarizes the neuronal models based on the Hodgkin and Huxley (HH) formalism first described in the 1950s. The HH formalism is a deterministic nonlinear four-dimensional model of the membrane potential dynamics and it can easily be simulated with current computational resources. However, in order to make this approach tractable for mathematical analysis, it must be further simplified. One simplification is to linearize it about a fixed point, e.g., using the resting potential to model subthreshold neuronal behavior. Another common simplification is to reduce the four dimensions into one or two dimensions. In the context of the simplification of the HH model, we employ equivalent electronic circuits. The simplified approach is also related to the integrate-and-fire (IF) modeling approach, first proposed by Lapicque more than a century ago, and neuronal resonance behavior. This chapter includes several MATLAB® routines to simulate the different cellular models and includes instructions on how to build an electronic version of the IF model.

In this chapter we continue the analysis of the filters presented in Chapters 15 and 16 Chapter 1... more In this chapter we continue the analysis of the filters presented in Chapters 15 and 16 Chapter 15 Chapter 16 . Here we consider the filter in the broader context of a linear time-invariant system that was introduced in Chapter 13 . We relate the filter's properties in the time domain and frequency domain. We quantify the relationship between the filter's time constant and the so-called −3-dB point that indicates the transition between pass band and stop band in the filter's frequency response. The filter's frequency characteristic is depicted in two types of plots: the Bode plot and the Nyquist plot. MATLAB ® scripts present examples of the analysis of the filter characteristic and the approach is employed to describe the characteristic measured in Chapter 15 . In addition to the use of sinusoidal or step inputs to perturb the filter, we present and demonstrate the approach where noise is employed to find the filter's frequency response.

Abstract In this chapter we review ordinary differential equations (ODEs) as a tool to model dyna... more Abstract In this chapter we review ordinary differential equations (ODEs) as a tool to model dynamics. We present examples of how to formulate them based on the dynamical system that needs to be modeled, and demonstrate the mathematical techniques one can employ to solve the equation analytically. We show how to solve linear differential equations with and without a forcing term, the so-called inhomogeneous and homogeneous ODEs, respectively. To illustrate the analysis of these equations, ODEs with first-order derivatives (e.g., d c / d t ) and second-order derivatives (e.g., d 2 c / d t 2 ) are used in the examples. Next, we show how higher-order ODEs can be represented as a set of first-order ones, and how this leads to a formalism in matrix/vector notation that can be efficiently analyzed using techniques from linear algebra. To complete the overview of the available tools for solving ODEs, the final part of this chapter briefly refers to application of Laplace and Fourier transforms (see also Chapter 12 ) to solve them.

We detect seizures in newborn infants using a novel method derived from triple correlation, which... more We detect seizures in newborn infants using a novel method derived from triple correlation, which integrates spatial and temporal structure in neonatal electroencephalograms (EEGs). Triple correlation natively encompasses analogues to a variety of lower-order approaches (auto-correlation, cross-correlation) in addition to introducing higher-order signals, so we hypothesized that our approach would both effectively detect and differentiate notoriously difficult-to-detect and heterogeneous neonatal seizures. Indeed, our method in its simplest form performs comparably well to a current standard of care, amplitude-integrated EEG (aEEG), and by some measures outperforms aEEG, suggesting at a minimum that a combination of triple correlation and aEEG could produce a more effective first-line bedside detector. Moreover, we find that the triple correlation seizure-signal varies between patients, with 1) differences in dominance of either within or between channel correlations and 2) differin...

Signal Processing for Neuroscientists, 2018

This chapter discusses the applications of the Fourier transform in spectral analysis and medical... more This chapter discusses the applications of the Fourier transform in spectral analysis and medical imaging. The raw complex-valued output of a fast Fourier transform or discrete Fourier transform is difficult to interpret directly. The most common approach is to present the power spectrum of a given signal. A related approach to displaying the results of spectral analysis is to show amplitude or phase spectra. Spectral analysis is often used in electroencephalography (EEG) analysis to evaluate the classical EEG frequency bands. In physiological signals, interpretation of spectra requires caution because these time series are rarely stationary and usually contain nonperiodic and nonsinusoidal components. The chapter shows examples of the Radon transform, the Fourier slice theorem, and filtered backprojection as each applies to CT image reconstruction. The use of the 2-D Fourier transform and k space are described in the context of magnetic resonance imaging (MRI) application. These techniques require reconstruction of a density function representing the internal structure of an object from sensor readings taken from outside that object.

Signal Processing for Neuroscientists, 2018

Abstract This chapter provides an overview of different types of models for studying the activity... more Abstract This chapter provides an overview of different types of models for studying the activity of nerve cells. It summarizes the neuronal models based on the Hodgkin and Huxley (HH) formalism first described in the 1950s. The HH formalism is a deterministic nonlinear four-dimensional model of the membrane potential dynamics and it can easily be simulated with current computational resources. However, in order to make this approach tractable for mathematical analysis, it must be further simplified. One simplification is to linearize it about a fixed point, e.g., using the resting potential to model subthreshold neuronal behavior. Another common simplification is to reduce the four dimensions into one or two dimensions. In the context of the simplification of the HH model, we employ equivalent electronic circuits. The simplified approach is also related to the integrate-and-fire (IF) modeling approach, first proposed by Lapicque more than a century ago, and neuronal resonance behavior. This chapter includes several MATLAB® routines to simulate the different cellular models and includes instructions on how to build an electronic version of the IF model.

Non UBCUnreviewedAuthor affiliation: The University of ChicagoFacult

Clinical Neurophysiology, 2018

Signal Processing for Neuroscientists, 2010

Gaussian white noise (GWN) allows one to create a series with orthogonal terms that can be estima... more Gaussian white noise (GWN) allows one to create a series with orthogonal terms that can be estimated sequentially with the Lee–Schetzen cross-correlation method. This approach can be adapted when the system's natural input consists of impulse trains such as a spike train. Identifying a system with an impulse train as input is discussed in this chapter. This approach was described by Krausz (1975). Poisson–Wiener operators are orthogonal to all lower-order Volterra operators, which is analogous to the development of the Wiener series with a GWN input. For the generation of a series of random numbers following a Gaussian or a uniform distribution, MATLAB commands randn and rand are used, respectively. A standard MATLAB command for generating a series of intervals according to a Poisson process does not exist. Therefore a Monte Carlo technique is applied to create such an impulse train according to a Poisson process.

Signal Processing for Neuroscientists, 2010

This first chapter describes two of the advanced techniques in signal processing: Lomb's algo... more This first chapter describes two of the advanced techniques in signal processing: Lomb's algorithm and the Hilbert transform. The astrophysicist Lomb developed an algorithm for spectral analysis to deal with signals consisting of unevenly sampled data. The idea of Lomb's algorithm is similar to the development of the Fourier series, namely, to represent a signal by a sum of sinusoidal waves. The Hilbert transform allows one to compute the instantaneous phase and amplitude of a signal. The fact that one can determine these two metrics in an instantaneous fashion is unique because usually this type of parameter can only be associated with an interval of the signal. For example, in spectral analysis the spectrum is computed for an epoch and the spectral resolution is determined by epoch length. Being able to determine parameters such as the phase instantaneously is especially useful if one wants to determine relationships between multiple signals generated within a neuronal network.

Gustatory sensitivity of larvae belonging to nine different Yponomeuta species was studied. A con... more Gustatory sensitivity of larvae belonging to nine different Yponomeuta species was studied. A conspicuous behavioural difference between these closely related species is represented in their host plant preferences. Electrophysiological and morphological examination of the anteriorly-located sensilla revealed that a restricted number of contact chemoreceptors are present. A tip recording technique was used to record gustatory responses of neurones in the lateral and medial styloconic sensilla. The following chemicals were applied as stimuli: sucrose, sedoheptulose, sorbitol, dulcitol, coumarin, phloridzin, salicin, prunasin, rutin, (+)-catechin, the trisodium salt of isocitric acid, ZnCl 2 , NaCl, and KCl. Most of these stimuli were selected on the basis of chemical composition of the array of host plants of the larvae studied (Table 1). Different spike amplitudes displayed in the recordings indicated that most constituents tested are perceived by single cells. Ionized substances sometimes excited more than one cell. In several recordings, a delayed response appeared. This phenomenon seems related to variation in diameter of the distal pore by which inward diffusion of the stimulus takes place. Analysis of the time course of the recordings indicated that the effect of diffusion of stimulus molecules on the initial neural response may be considerable for the non-ionized substances tested. The functional diversity displayed in the interspecific sensitivity patterns can be partly explained by composition of the host plants on which the larvae feed (Table 2). Compounds of one type act as phagostimulants, and some of these compounds are sufficiently host-specific to act as host-recognition factors. A second type of chemical signal inhibits "wrong" food intake; these substances are classified as deterrents. A few gustatory responses appeared non-adaptive, this type of sensitivity is discussed from an evolutionary point of view. 7the maxilla or the maxillary palpi leads to acceptance of normally rejected plants (Torii and Morii 1948; Dethier 1953; Waldbauer and Fraenkel 1961). Assuming a relationship between speciation and host plant choice, the latter studies indicate the significance of the gustatory organs from an evolutionary point of view. Various theories on sympatric and allopatric speciation of phytophagous insects have been described by Dethier (1952), Bush (1974), Wiebes (1976) and Labeyrie (1978). In recent comparative studies on neural and behaviouiral gustatory responses of larvae belonging to the genus Yponomeuta it is suggested that there exists a persistence of sensory sensitivities to constituents of the ancestral host plant (Gerrits-Heybroek et al., 1978; van Drongelen 1978). Because many species of the genus Yponomeuta seem closely related, but show different food regimes (Friese 1960; Herrebout et al., 1976), they are good candidates for a comparative study on insect-host plant relationship. This paper aims to describe gustatory activity during application of plant constituents, as recorded in the lateral and medial styloconic sensilla of nine Yponomeuta species. Parameters of the neural responses are considered in connection with primary events of the peripheral perception mechanism. To evaluate aspects of the relationship between chemosensitivity and occurrence of plant constituents in the host, an attempt is made to state expectations concerning the sensitivity patterns. The distribution of various compounds over the host plants and assumptions on neural mechanisms represent the input of these expectations. Deviations from expected sensitivity are discussed from an evolutionary point of view. A short description of a part of this work appeared previously (van Drongelen 1978) and preliminary data on the sugar receptors have been described by Schoonhoven et al. (1977). Methods Animals. Most larvae of the Yponomeuta species (Lepidoptera, Yponomeutidae) were collected in the field. Larvae of Y. vigintipunctatus and Y. irrorellus were reared in the laboratory on their host plant. The animals were stored in a refrigerator at 6°C for a maximal duration of one month. A list of the Yponomeuta species studied and their associated hosts is shown in the first two rows of Table 1. It can be seen that Y. padellus is oligophagous and the remaining species are monophagous. In this paper, the host-race of Y. padellus feeding on Crataegus species will be referred to as host-race 1 and the one feeding on Prunus spinosa as host-race 2. It was found practical to use the host plant on which the larvae were collected as the main criterion for species identification. On morphological grounds larvae of Y. vigintipunctatus and of Y. plumbellus can be clearly distinguished from those of the other species studied; the first two species belong to a different taxonomie group (group B) than the remaining ones (group A) (Gerrits-Heybroek et al-., 1978). Larvae of Y. cagnagellus were distinguished from those of Y. irrorellus by morphological identification of adult individuals of the latter species, which were reared in a laboratory culture. Morphology. The mouth parts of the insects to be investigated were fixed in Bouin. The fixed preparations were placed in freon liquid, subsequently in liquid nitrogen and were freeze dried. Dried preparations were dissected and coated with gold. The sensory organs were examined in a scanning electron microscope (Jeol, JSM-Ug) in the Service Institute for Technical Physics in Agriculture in Wageningen. Ent. exp. & appl., (in press) CHAPTER 3 Behavioural responses of two small ermine moth species (Lepidoptera: Yponomeutidae) to plant constituents

Chemistry and Physics of Lipids - CHEM PHYS LIPIDS, 2002

Seizure cancellation in a network model with Hodgkin-Huxley type elements was evaluated by applyi... more Seizure cancellation in a network model with Hodgkin-Huxley type elements was evaluated by applying different stimulation paradigms. The simulated activities of single neurons and a computed field potential during different states of the network were the basis of this evaluation. Three types of electrostimulation were investigated: hyperpolarization of the neurons, depolarization, and an electric stimulus that is proportional to the inverted field potential. We demonstrated that a seizure, represented by synchronized network activity, and consequently a large amplitude field potential, could be effectively stopped by the third type of stimulus. The depolarization and hyperpolarization paradigms failed to desynchronize the activities.

Solving differential equations analytically is not always the easiest strategy, or even possible.... more Solving differential equations analytically is not always the easiest strategy, or even possible. In these cases one may use a numerical solution to characterize the dynamics governed by the ordinary differential equation (ODE). A diagram of the flow plotted versus time can create insight into the dynamical behavior governed by the ODE. In addition, the so-called phase space representation, where the dynamics is represented by a vector at each state of the system, is a helpful tool to qualitatively assess system dynamics. Here, we examine both numerical solutions and graphical representations of the dynamics using two growth models: one model with unrestricted growth (a linear ODE), and one where the growth is restricted by limited resources (the logistic equation, a nonlinear ODE). We use different techniques and MATLAB ® examples to demonstrate the numerical approach: Euler's method, the improved Euler's method, and the fourth-order Runge–Kutta algorithm. In the final part we briefly discuss methods for obtaining solutions for partial differential equations.

The Journal of Physiology, Apr 1, 1978

1. Receptor cell activity in the frog's eminentia olfactoria was recorded using m... more 1. Receptor cell activity in the frog's eminentia olfactoria was recorded using metal-filled micro-electrodes. 2. Several units discharged spontaneously with a mean frequency lower than 0.2 spikes per sec, or were silent in periods of up to 5 min. The other units displayed spontaneous activities between 0.2 and 1.05 spikes per sec; their activity could be modelled with a Poisson process. 3. Near-threshold responses to odour stimulation were investigated, considering several stimulations within a small concentration range. Low concentration stimulations were sometimes followed by a response, sometimes not. The concept of response probability is introduced to describe this incertitude. 4. The distribution of the number of spikes in several odour trials at low concentrations showed a reasonable agreement with two types of Poisson distribution. 5. The findings are discussed in connexion with receptor cell sensitivity and the excitation of second order neurones in the bulb.

Journal of Theoretical Biology, Mar 7, 1978

ABSTRACT