Browsing by Author "Kemere, Caleb T"

Now showing 1 - 5 of 5
  • Thumbnail Image
    Item
    Analysis and Application of a Realtime Closed-loop Hippocampal Sharp-wave Ripple Disruption System
    (2019-08-09) Dutta, Shayok; Kemere, Caleb T
    Closed-loop modulation of neural activity facilitates causal tests of hypothesis to link temporal neural circuit functions to behavior and cognition. One such dynamically-changing neural circuit in the brain is the hippocampus, responsible for learning and memory. Transient bursts of neural activity in the hippocampus during periods of quiescence, sharp-wave ripple (SWR) events --- ≈80--150 ms periods comprising large 150--250 Hz oscillations --- have previously been established to serve a critical role in memory consolidation and recall through selective closed-loop interactions. The timescale within which these events occur indicate that latency in systems used for closed-loop perturbations is of significant consequence when compared to other longer-lasting events. However, though this approach is beginning to become widespread, performance trade-offs involved in the fabrication of SWR detection and disruptions have not been fully characterized, thereby, limiting the understanding of results using this experimental procedure. In this thesis, I present the development and characterization of a low-latency closed-loop system for SWR detection and modulation engineered to interface with two separate widely-used neural data acquisition hardware platforms. The analysis and characterization of the system is done with a two-pronged approach: (1) generation of a synthetic data to model SWR activity in order explore the parameter space of the realtime detection algorithm and (2) quantification of the realtime in vivo performance of the algorithm and system via offline simulation. Lastly, realtime in vivo detections of SWRs are shown match the simulated detections with added hardware data transmission latency. By performing this characterization, I have established a framework for analyzing other closed-loop systems for neural perturbations. Next, I deploy this system in a neuroscientific closed-loop experiment. In particular, rodents have an innate curiosity to explore novel contexts and objects resulting in the spending more time with novel objects than familiar ones. Work has been done through targeted lesion studies to establish hippocampal dependence of this object recognition memory via novel object test (NOT) paradigms. More recent work has explored correlations of hippocampal CA1 signatures, such as fast-gamma oscillations, to be of importance in NOT object-place recognition memories and demonstrated predictable changes in SWRs. More specifically, SWRs have been shown to increase after encoding of novel objects (or familiar objects in novel locations) and general novelty in the test; however, the concomitant spiking activity has not been correlated with object-place pairings leading to questioning the role of these events in object-place recognition memories. By using the previously engineered system, I demonstrate preliminary results showing suppression of this novelty preference in a rat during a hippocampally dependent novel object test paradigm. This preliminary finding suggests SWRs play a role in object-place recognition memory and not just in sequential goal oriented experimental paradigms as previous results have shown.
  • Thumbnail Image
    Item
    Efficient Software and Algorithms for the Representation and Analysis of Neural Data
    (2020-06-12) Chu, Joshua; Kemere, Caleb T
    Recent technological advances have enabled neural recordings consisting of hundreds to thousands of channels. As the pace of these developments continues to grow rapidly, it is imperative to have fast, flexible tools supporting the analysis of neural data gathered by such large scale modalities. Here we introduce the nelpy (neuroelectrophysiology in Python) ecosystem, a software suite that enables the neuroscientist and engineer to represent data in a containerized manner and perform typical operations on those objects. The base nelpy package supports a variety of common experimental types encountered by electrophysiologists, including sampled continuous signals, spike trains, and binned spike trains, all defined over an arbitrary domain. Additional features include a highly customizable plotting library for rapid data exploration and visualization. The essential nelpy functionality of data representation can be augmented by other modules that perform a variety of analyses. As an example of extensibility, we focus on signal processing and spectral analyses provided by ghostipy (grand harmonization of spectral techniques in Python). Besides providing functionality such as optimal digital filters and time-frequency transforms, ghostipy implements analyses that outperform commercial software in both time and space complexity for high channel count data. Overall, our software suite reduces frequently encountered bottlenecks in the experimental pipeline, and we believe this toolset will enhance both the portability and scalability of neural data analysis.
  • Thumbnail Image
    ItemEmbargo
    Interrogating the role of hippocampal sharp-wave ripples in spontaneous learning
    (2023-10-26) Dutta, Shayok; Kemere, Caleb T
    Rodents naturally explore novelty in contexts, objects, or locations, showing a preference for the unfamiliar. Lesion studies establish this novelty preference as dependent on the hippocampus, particularly through object recognition memory (ORM) paradigms. Recent research links hippocampal CA1 LFP signatures, including fast-gamma and beta-band oscillations, to object-place recognition memory. This work observes predictable increases in transient hippocampal events like sharp-wave ripples (SWRs). Although SWRs increase after encoding novelty, the associated spiking activity does not necessarily align with object-place pairings, raising questions about the role of SWRs in curiosity-driven spontaneous learning tasks. This thesis employs selective modulation of SWR activity using a previously engineered open-source, closed-loop SWR detection system in an ORM displaced object paradigm. Results reveal that suppressing SWR activity during object encoding and post-encoding rest sessions significantly impairs object-place recognition memory. Analysis of recorded CA1 LFP data shows statistical changes in SWR rates between disruption and control groups, while preserving general exploratory behavior across these groups. Further analysis correlates changes in SWR duration from pre-encoding to post-encoding rest sessions with discrimination measures. These correlations suggest that longer ripple durations lead to higher novelty preference scores. To investigate this phenomenon, the study employs a novel algorithm for the selective interrogation of longer-duration ripples during post-encoding rest sessions. In conclusion, the findings indicate that SWRs, particularly longer-duration ripples, critically influence object-place recognition memory driven by curiosity. This work advances our understanding of memory consolidation and the processing of spontaneous memories, extending beyond the traditionally studied realm of food or reward-driven spatial memories.
  • Thumbnail Image
    Item
    Methods for Ripple Detection and Spike Sorting During Hippocampal Replay
    (2015-09-21) Sethi, Ankit; Kemere, Caleb T; Aazhang, Behnaam; Robinson, Jacob
    In the rat hippocampus, fast oscillations termed sharp wave ripples and an associated sequential firing of neurons, termed replay, have been identified as playing a crucial role in memory formation and learning. The term 'replay' is used since the observed spiking encodes patterns of past experiences. To determine the role of replay in learning and decision making, a need arises for systems that can decode replay activity observed during ripples. This necessitates online algorithms for both spike sorting and ripple detection at low latencies. In my work, I have developed and tested an improved method for ripple detection and tested its performance against previous methods. Further, I have optimized a recently proposed spike sorting algorithm based on real-time bayesian inference so that it can run online in a multi-tetrode scenario, and implemented it, along with ripple detection, for the open-source electrophysiological suite, "open-ephys". The algorithm's parameters were also analyzed for their suitability in operating in an unsupervised scenario. These two modules are integrated to form a system uniquely suited to decoding neuronal sequences during sharp wave ripple events.
  • Thumbnail Image
    Item
    Optimizing extra-cranial light source location based on light intensity simulation in hippocampus
    (2024-07-24) del Rio Pulido, Daniela; Kemere, Caleb T
    The hippocampus (HC) is a deep brain region critical for long-term memory formation. Since the early 2000s, neuroscience has changed from using temporal and spatial non-specific tools to relying on optogenetics because of its selective control over neural activity using light. Recent studies have delivered light to deep brain regions through implanted optical fibers; however, this invasive procedure inevitably causes tissue damage. Neuroscience research could improve by combining temporal and spatial specificity provided by optogenetics and non-invasive light delivery. In this work, I simulate light intensity on the surface of the CA1 layer of HC using a 2-layer simulation of the Lambert-Beer equation. Anatomical brain regions were determined using the open-source Waxholm Space atlas of the Sprague Dawley rat brain. Given a fixed number of light sources, we found the optimal location of the light source(s) on a rat's skull to maximize light intensity on the surface of CA1, the target region. The real-world application of placing light sources in optimal skull locations will enable minimal brain damage while preserving the temporal and spatial specificity provided by optogenetics.