Single and Multiple Change Point Detection in Spike Trains: Comparison of Different CUSUM Methods

Detecting change points in neural population activity with contrastive metric learning

Finding points in time where the distribution of neural responses changes (change points) is an important step in many neural data analysis pipelines. However, in complex and free behaviors, where we see different types of shifts occurring at different rates, it can be difficult to use existing methods for change point (CP) detection because they can't necessarily handle different types of changes that may occur in the underlying neural distribution. Additionally, response changes are often sparse in high dimensional neural recordings, which can make existing methods detect spurious changes. In this work, we introduce a new approach for finding changes in neural population states across diverse activities and arousal states occurring in free behavior. Our model follows a contrastive learning approach: we learn a metric for CP detection based on maximizing the Sinkhorn divergences of neuron firing rates across two sides of a labeled CP. We apply this method to a 12-hour neural recording of a freely behaving mouse to detect changes in sleep stages and behavior. We show that when we learn a metric, we can better detect change points and also yield insights into which neurons and sub-groups are important for detecting certain types of switches that occur in the brain.

Multiple change point estimation of trends in Covid-19 infections and deaths in India as compared with WHO regions

The present study aims at estimating the multiple change points for the time series data of COVID-19 confirmed cases and deaths and trend estimation within the estimated multiple change points (MCP) in India as compared with WHO regions. The data were described using descriptive statistical measures, and for the estimation of change point's E-divisive procedure was employed. Further, the trend within the estimated change points was tested using Sen's slope and Mann Kendal tests. India, along with the African Region, American region, and South East Asia regions experienced a significant surge in the fresh cases up to the 5th Change point. Among the WHO regions, The American region was the worst hit by the pandemic in case of fresh cases and deaths. While the European region experienced an early negative trend of fresh cases during the 3rd and 4th change point, but later the situation reversed by the 5th (7th July 2020) and 6th (6th August 2020) change point. The trend of deaths in India and the South-East Asia Region was similar, and global deaths had a negative trend from the 4th (17th May 2020) Change point onwards. The change points were estimated with prefixed significance level α < 0.002. Infections and deaths were positively significant for India and SEARO region across change points. Infection was significant at every 30 days interval across other WHO regions, and any delay in the infections was due to the interventions. The European region is expected to have a second wave of positive infections during the 5th and 6th change points though the early two change points were negatively significant. The study highlights the efficacy of change point analysis in understanding the dynamics of covid-19 cases in India and across the world. It further helps to develop effective public health strategies.

Neural Circuits for Dynamics-Based Segmentation of Time Series

The brain must extract behaviorally relevant latent variables from the signals streamed by the sensory organs. Such latent variables are often encoded in the dynamics that generated the signal rather than in the specific realization of the waveform. Therefore, one problem faced by the brain is to segment time series based on underlying dynamics. We present two algorithms for performing this segmentation task that are biologically plausible, which we define as acting in a streaming setting and all learning rules being local. One algorithm is model based and can be derived from an optimization problem involving a mixture of autoregressive processes. This algorithm relies on feedback in the form of a prediction error and can also be used for forecasting future samples. In some brain regions, such as the retina, the feedback connections necessary to use the prediction error for learning are absent. For this case, we propose a second, model-free algorithm that uses a running estimate of the autocorrelation structure of the signal to perform the segmentation. We show that both algorithms do well when tasked with segmenting signals drawn from autoregressive models with piecewise-constant parameters. In particular, the segmentation accuracy is similar to that obtained from oracle-like methods in which the ground-truth parameters of the autoregressive models are known. We also test our methods on data sets generated by alternating snippets of voice recordings. We provide implementations of our algorithms at https://github.com/ttesileanu/bio-time-series.

Cortical representation of group social communication in bats

Social interactions occur in group settings and are mediated by communication signals that are exchanged between individuals, often using vocalizations. The neural representation of group social communication remains largely unexplored. We conducted simultaneous wireless electrophysiological recordings from the frontal cortices of groups of Egyptian fruit bats engaged in both spontaneous and task-induced vocal interactions. We found that the activity of single neurons distinguished between vocalizations produced by self and by others, as well as among specific individuals. Coordinated neural activity among group members exhibited stable bidirectional interbrain correlation patterns specific to spontaneous communicative interactions. Tracking social and spatial arrangements within a group revealed a relationship between social preferences and intra- and interbrain activity patterns. Combined, these findings reveal a dedicated neural repertoire for group social communication within and across the brains of freely communicating groups of bats.

Decoding pain from brain activity

Pain is a dynamic, complex and multidimensional experience. The identification of pain from brain activity as neural readout may effectively provide a neural code for pain, and further provide useful information for pain diagnosis and treatment. Advances in neuroimaging and large-scale electrophysiology have enabled us to examine neural activity with improved spatial and temporal resolution, providing opportunities to decode pain in humans and freely behaving animals. This topical review provides a systematical overview of state-of-the-art methods for decoding pain from brain signals, with special emphasis on electrophysiological and neuroimaging modalities. We show how pain decoding analyses can help pain diagnosis and discovery of neurobiomarkers for chronic pain. Finally, we discuss the challenges in the research field and point to several important future research directions.

Online Detection of Multiple Stimulus Changes Based on Single Neuron Interspike Intervals

Nervous systems need to detect stimulus changes based on their neuronal responses without using any additional information on the number, times, and types of stimulus changes. Here, two relatively simple, biologically realistic change point detection methods are compared with two common analysis methods. The four methods are applied to intra- and extracellularly recorded responses of a single cricket interneuron (AN2) to acoustic simulation. Solely based on these recorded responses, the methods should detect an unknown number of different types of sound intensity in- and decreases shortly after their occurrences. For this task, the methods rely on calculating an adjusting interspike interval (ISI). Both simple methods try to separate responses to intensity in- or decreases from activity during constant stimulation. The Pure-ISI method performs this task based on the distribution of the ISI, while the ISI-Ratio method uses the ratio of actual and previous ISI. These methods are compared to the frequently used Moving-Average method, which calculates mean and standard deviation of the instantaneous spike rate in a moving interval. Additionally, a classification method provides the upper limit of the change point detection performance that can be expected for the cricket interneuron responses. The classification learns the statistical properties of the actual and previous ISI during stimulus changes and constant stimulation from a training data set. The main results are: (1) The Moving-Average method requires a stable activity in a long interval to estimate the previous activity, which was not always given in our data set. (2) The Pure-ISI method can reliably detect stimulus intensity increases when the neuron bursts, but it fails to identify intensity decreases. (3) The ISI-Ratio method detects stimulus in- and decreases well, if the spike train is not too noisy. (4) The classification method shows good performance for the detection of stimulus in- and decreases. But due to the statistical learning, this method tends to confuse responses to constant stimulation with responses triggered by a stimulus change. Our results suggest that stimulus change detection does not require computationally costly mechanisms. Simple nervous systems like the cricket's could effectively apply ISI-Ratios to solve this fundamental task.

Honey bees flexibly use two navigational memories when updating dance distance information

Honey bees can communicate navigational information which makes them unique amongst all prominent insect navigators. Returning foragers recruit nest mates to a food source by communicating flight distance and direction using a small scale walking pattern: the waggle dance. It is still unclear how bees transpose flight information to generate corresponding dance information. In single feeder shift experiments, we monitored for the first time how individual bees update dance duration after a shift of feeder distance. Interestingly, the majority of bees (86%) needed two or more foraging trips to update dance duration. This finding demonstrates that transposing flight navigation information to dance information is not a reflexive behavior. Furthermore, many bees showed intermediate dance durations during the update process, indicating that honey bees highly likely use two memories: (i) a recently acquired navigation experience and (ii) a previously stored flight experience. Double-shift experiments, in which the feeder was moved forward and backward, created an experimental condition in which honey bee foragers did not update dance duration; suggesting the involvement of more complex memory processes. Our behavioral paradigm allows the dissociation of foraging and dance activity and opens the possibility of studying the molecular and neural processes underlying the waggle dance behavior.

Ensembles of change-point detectors: implications for real-time BMI applications

Brain-machine interfaces (BMIs) have been widely used to study basic and translational neuroscience questions. In real-time closed-loop neuroscience experiments, many practical issues arise, such as trial-by-trial variability, and spike sorting noise or multi-unit activity. In this paper, we propose a new framework for change-point detection based on ensembles of independent detectors in the context of BMI application for detecting acute pain signals. Motivated from ensemble learning, our proposed "ensembles of change-point detectors" (ECPDs) integrate multiple decisions from independent detectors, which may be derived based on data recorded from different trials, data recorded from different brain regions, data of different modalities, or models derived from different learning methods. By integrating multiple sources of information, the ECPDs aim to improve detection accuracy (in terms of true positive and true negative rates) and achieve an optimal trade-off of sensitivity and specificity. We validate our method using computer simulations and experimental recordings from freely behaving rats. Our results have shown superior and robust performance of ECPDS in detecting the onset of acute pain signals based on neuronal population spike activity (or combined with local field potentials) recorded from single or multiple brain regions.

A Novel Covert Agent for Stealthy Attacks on Industrial Control Systems Using Least Squares Support Vector Regression

Research on stealthiness has become an important topic in the field of data integrity (DI) attacks. To construct stealthy DI attacks, a common assumption in most related studies is that attackers have prior model knowledge of physical systems. In this paper, such assumption is relaxed and a covert agent is proposed based on the least squares support vector regression (LSSVR). By estimating a plant model from control and sensory data, the LSSVR-based covert agent can closely imitate the behavior of the physical plant. Then, the covert agent is used to construct a covert loop, which can keep the controller’s input and output both stealthy over a finite time window. Experiments have been carried out to show the effectiveness of the proposed method.

Real-time particle filtering and smoothing algorithms for detecting abrupt changes in neural ensemble spike activity

Sequential change-point detection from time series data is a common problem in many neuroscience applications, such as seizure detection, anomaly detection, and pain detection. In our previous work (Chen Z, Zhang Q, Tong AP, Manders TR, Wang J. J Neural Eng 14: 036023, 2017), we developed a latent state-space model, known as the Poisson linear dynamical system, for detecting abrupt changes in neuronal ensemble spike activity. In online brain-machine interface (BMI) applications, a recursive filtering algorithm is used to track the changes in the latent variable. However, previous methods have been restricted to Gaussian dynamical noise and have used Gaussian approximation for the Poisson likelihood. To improve the detection speed, we introduce non-Gaussian dynamical noise for modeling a stochastic jump process in the latent state space. To efficiently estimate the state posterior that accommodates non-Gaussian noise and non-Gaussian likelihood, we propose particle filtering and smoothing algorithms for the change-point detection problem. To speed up the computation, we implement the proposed particle filtering algorithms using advanced graphics processing unit computing technology. We validate our algorithms, using both computer simulations and experimental data for acute pain detection. Finally, we discuss several important practical issues in the context of real-time closed-loop BMI applications. NEW & NOTEWORTHY Sequential change-point detection is an important problem in closed-loop neuroscience experiments. This study proposes novel sequential Monte Carlo methods to quickly detect the onset and offset of a stochastic jump process that drives the population spike activity. This new approach is robust with respect to spike sorting noise and varying levels of signal-to-noise ratio. The GPU implementation of the computational algorithm allows for parallel processing in real time.