ISI Sequences Research Articles

Analysis of noise-induced temporal correlations in neuronal spike sequences

We investigate temporal correlations in sequences of noise-induced neuronal spikes, using a symbolic method of time-series analysis. We focus on the sequence of time-intervals between consecutive spikes (inter-spike-intervals, ISIs). The analysis method, known as ordinal analysis, transforms the ISI sequence into a sequence of ordinal patterns (OPs), which are defined in terms of the relative ordering of consecutive ISIs. The ISI sequences are obtained from extensive simulations of two neuron models (FitzHugh-Nagumo, FHN, and integrate-and-fire, IF), with correlated noise. We find that, as the noise strength increases, temporal order gradually emerges, revealed by the existence of more frequent ordinal patterns in the ISI sequence. While in the FHN model the most frequent OP depends on the noise strength, in the IF model it is independent of the noise strength. In both models, the correlation time of the noise affects the OP probabilities but does not modify the most probable pattern.

Interspike intervals as a discrete time series with history and randomness

Neurons are fundamental components of neural systems. Organized neural activity is key for proper function at many scales, from small central pattern generating circuits to large populations of neurons. Neurons can show considerable variability in their spiking regularity; this variability is attributed to a number of different sources, including probabilistic synaptic processes and stochastic ion channel activity [1]. The role of variability in neural processes is complex: for example, more variability in an input signal may enhance the consistency of output spiking [2], but spike variability poses problems for systems or theories that require precise event timing, such as spike-timing dependent plasticity [3]. In this work, we characterized the variability of intrinsically active neurons over time and found that all cells studied shared certain features. We measured intracellularly from synaptically-isolated, endogenously spiking Aplysia californica invertebrate neurons for 2 or more hours in a temperature-controlled preparation to determine the regularity of interspike intervals. We determined that sequential ISIs show strong positive correlation with prior ISIs and this correlation remains high for many lag values, indicating that the ISI sequence shows history. In contrast, sequential values of the change in ISI (delta ISI) show a single prominent negative autocorrelation coefficient; this indicates that delta ISI correlates strongest with one previous delta ISI, and the negative sign suggests a tendency for delta ISI values to remain within some range. We find that ISI sequences from intrinsically spiking neurons in this system can be represented using autoregressive integrated moving average (ARIMA) models. In addition, delta ISI distributions are symmetric and Gaussian-like with heavy tails. Recent literature reports that heavy-tailed ISI distributions that show positive serial correlation values may result from stochastic activity in populations of adapting ion channels [4,5]; we are currently investigating the cellular features that can produce an ARIMA process, paying particular attention to channel dynamics. We present here an ISI variability model that captures statistical features of experimental data, which are essentially history and randomness. Existing explanations for neural spiking variability include properties of synaptic conductances [6], neural input [7], and ion channels [8]; these factors are modeled in simulation as or in combination with random processes that represent noisy inputs, which are then added to the membrane voltage. The phenomenological description of the ISI sequence itself as an ARIMA process with stochastic noise and history is a unique addition to the collection of neural noise models; it does not require explicit description or addition of cellular-level processes, it constrains correlational features rather than ISI distribution shape, and it represents intrinsic spike variability that is not due to synapses. This work was funded by NIH NS 54281.

Fire patterns of modified HH neuron under external sinusoidal ELF stimulus

Abstract Neuron as the main information carrier in neural systems is able to generate diverse fire trains in response to different stimuli. In this paper, the stimulus frequency is taken as the bifurcation parameter, and ISI is considered to be one of the state variables. Via numerical simulation, we mainly concentrate on the kinds of fire patterns that the modified HH neuron model displays such as period- n , bursting, and modulation fire patterns, etc. under the effect of external sinusoidal ELF electric field, and the relation between the ISI sequences and the external stimulus just like synchronization and transition in the manner of pitchfork bifurcation. In addition, an explanation is put forwards from the electrophysiology point of view to try to interpret why neurons generate so many different kinds of ISI sequences.

Molecular characterization and genomic distribution of Isis: a new retrotransposon of Drosophila buzzatii

A new transposable element, Isis, is identified as a LTR retrotransposon in Drosophila buzzatii. DNA sequence analysis shows that Isis contains three long ORFs similar to gag, pol and env genes of retroviruses. The ORF1 exhibits sequence homology to matrix, capsid and nucleocapsid gag proteins and ORF2 encodes a putative protease (PR), a reverse transcriptase (RT), an Rnase H (RH) and an integrase (IN) region. The analysis of a putative env product, encoded by the env ORF3, shows a degenerated protein containing several stop codons. The molecular study of the putative proteins coded by this new element shows striking similarities to both Ulysses and Osvaldo elements, two LTR retrotransposons, present in D. virilis and D. buzzatii, respectively. Comparisons of the predicted Isis RT to several known retrotransposons show strong phylogenetic relationships to gypsy-like elements, particulary to Ulysses retrotransposon. Studies of Isis chromosomal distribution show a strong hybridization signal in centromeric and pericentromeric regions, and a scattered distribution along all chromosomal arms. The existence of insertional polymorphisms between different strains and high molecular weight bands by Southern blot suggests the existence of full-sized copies that have been active recently. The presence of euchromatic insertion sites coincident between Isis and Osvaldo could indicate preferential insertion sites of Osvaldo element into Isis sequence or vice versa. Moreover, the presence of Isis in different species of the buzzatii complex indicates the ancient origin of this element.

Implementation and evaluation of frequency offset corrected inversion (FOCI) pulses on a clinical MR system.

FOCI pulses are a variant of hyperbolic secant pulses in which the RF amplitude, RF frequency, and gradient waveform are all modulated by the same function A(t). This increases the usable gradient amplitude without requiring a corresponding increase in RF amplitude. In this paper the implementation and inversion profiles of FOCI pulses on a clinical MR system are described, showing improved slice definition and chemical-shift offset behavior. Their adiabatic behavior is confirmed, and their use with an ISIS sequence for localized MRS is illustrated. Finally the effects of waveform digitization are considered, and the implications for SAR and dB/dt are discussed.

Saturation correction in human cardiac 31P MR spectroscopy at 1.5 T

This study was conducted to verify the validity of using saturation factors obtained from unlocalized 31P spectra containing both chest wall and heart muscle signals for correcting human heart muscle phosphocreatine/beta-adenosine triphosphate (PCr/beta-ATP) ratios. Saturation factors and T1 relaxation times were determined from 31P magnetic resonance spectra of human chest wall and heart muscle simultaneously in healthy volunteers using one-dimensional spectroscopic imaging in combination with a two-dimensional ISIS sequence by using adiabatic 180 degrees inversion and adiabatic 90 degrees excitation pulses at 1.5 T. Blood corrected saturation factors for PCr/beta-ATP at a TR of 2.4 s were significantly different in heart muscle and chest wall muscle, 1.30 +/- 0.25 and 1.73 +/- 0.31, respectively (p < 0.05). T1 values for PCr and beta-ATP in heart muscle were 4.28 +/- 0.72 and 2.99 +/- 0.52 and in chest wall muscle 6.82 +/- 1.07 and 3.39 +/- 0.48, respectively. The T1(PCr)/T1(beta-ATP) ratios in chest wall and heart muscle were not identical. The mean PCr/beta-ATP ratios in heart and chest wall muscle of six healthy volunteers were 1.23 +/- 0.17 and 3.71 +/- 0.53, respectively.

Lokalisierte31P-NMR-Spektroskopie mit ISIS und Oberflächenspule:

A new method for image-guided localized phosphorus NMR spectroscopy of superficial tissues has been investigated using a 1.5 Tesla whole-body-MR-System. We used a surface coil combined with adiabatic excitation pulses and a modified ISIS sequence. This approach is related to imaging sequences and thus permits a flexible and accurate determination of the volume of interest from "conventional" proton images. The scope and advantages of the method are demonstrated by phantom studies. Clinical applications to the liver, renal transplants, and the mediastinum are described.

Experimental approaches to image localized human 31P NMR spectroscopy.

Experimental procedures for obtaining localized 31P NMR spectra of humans by means of the ISIS sequence are discussed in detail. The technique is optimized for use with volume coils and with surface coils in order to measure localized 31P NMR spectra of different tissues and organs. Selective frequency-modulated (FM) inversion and excitation pulses are applied for optimal inversion or excitation despite B1 inhomogeneity. Pulse imperfection may lead to spurious signal contributions from outside the selected volume; this contamination is reduced by using long pulse intervals, by properly ordering the ISIS acquisitions, and by using FM excitation pulses. Simultaneous measurement of multiple volumes was implemented by including an additional selective inversion pulse, and an extension of the ISIS addition/subtraction scheme. Localized T1 measurements with surface coils are implemented by using a B1-insensitive inversion pulse in the inversion recovery sequence. The quantitative reproducibility of localized 31P NMR spectra was verified. Absolute metabolite concentration can be determined after a suitable calibration of the 31P NMR spectrum. Localized shimming is required to obtain localized 31P NMR spectra of excellent spectral resolution. This is done by monitoring the 1H NMR signal from water by a single-shot localization technique. The techniques discussed can be applied to obtain spectra of brain, liver, heart, and other organs. 31P NMR spectra of intracranial tumors demonstrate its applicability in the examination of patients.

Imaging and localized spectroscopy of 13C by polarization transfer

The biological applications of 13C magnetic resonance spectroscopy are well known ( 1-5). Recently, there was evidence (6-9) suggesting that the citrate peak in the 13C spectrum may be useful as a marker for human prostate cancer. It was found in human prostate that the citrate concentration is small in the normal gland, negligible in ma lignant tumors (8)) and increased markedly in tissues with benign prostatic hypertrophy (BPH) (8, 9). Wh ile this evidence suggests citrate as possibly a specific marker for ma lignant prostatic tumor and BPH, it also underscores the problem in applying NMR spectroscopic techniques without localization under in vivo conditions where prostatic ma lignancy is often, if not always, accompanied by BPH. There are two possible solutions to the above problem. One is through mu ltidimensional chemical-shift imaging ( 10, I I ) and the other, localized spectroscopy. For carbon, either one of these two approaches presents many challenging problems. F irst, one must deal with the problem of low sensitivity and long T1. Second, because carbon has a low y, one needs four times as much gradient strength in imaging or localization of carbon as that of hydrogen. Third, since carbon has large chemical shifts, it is often difficult to separate the spatial from the spectral information. By using polarization transfer techniques, such as INEPT ( 12,13) or DEPT ( 14,15), the problems associated with low y nuclei, such as carbon, are converted into those of protons which are much easier to solve. Recently this idea was exploited by Norris et al. ( 16)) who applied a one-dimensional ISIS sequence ( 17) on the protons for the localization of carbon. In this Communication, we will demonstrate an alternative technique of carbon selection based on the idea ment ioned above and the conventional p lane excitation method commonly used in proton imaging. In this scheme, plane selection of carbon is obtained by selective proton enhancement. We will also demonstrate that one can gain a similar type of enhancement in carbon imaging by applying phase encoding on the proton, instead of the carbon, polarization. We shall call this an indirect spinwarp method. F igure 1 shows the method of localization of carbon by selective enhancement through excitation of hydrogen. The RF section of the pulse sequence in F ig. 1 is a DEPT sequence with a shaped proton 7r/2 pulse. Since the phase-cycl ing scheme of the DEPT sequence is so designed that only the proton-enhanced nuclei are observed

Amplification of the ArgF region in strain HfrP4X of E. coli K-12.

In E. coli K-12 the argF gene is flanked by ISI sequences in direct repeat. Mutants that overproduce the argF-coded enzyme ornithine transcarbamylase can be selected; we have shown that in one class of these mutants there is an approximately forty five-fold amplification of the region bounded by the ISI repeats. This class of mutants has been detected only in strains in which the F-factor is integrated in cis to the region.