An official website of the United States government
Optogenetics has transformed studies of neural circuit function, but remains challenging to apply in non-human primates (NHPs). A major challenge is delivering intense and spatially precise patterned photostimulation across large volumes in deep tissue. Here, we have developed and tested the Utah Optrode Array (UOA) to meet this critical need. The UOA is a 10×10 glass waveguide array bonded to an electrically-addressable μLED array. In vivo electrophysiology and immediate early gene (c-fos) immunohistochemistry demonstrate that the UOA allows for large-scale spatiotemporally precise neuromodulation of deep tissue in macaque primary visual cortex. Specifically, the UOA permits either focal (confined to single layers or columns), or large-scale (across multiple layers or columns) photostimulation of deep cortical layers, simply by varying the number of simultaneously activated μLEDs and/or the light irradiance. These results establish the UOA as a powerful tool for studying targeted neural populations within single or across multiple deep layers in complex NHP circuits.
more »
« less
An optrode array for spatiotemporally-precise large-scale optogenetic stimulation of deep cortical layers in non-human primates
Abstract Optogenetics has transformed studies of neural circuit function, but remains challenging to apply to non-human primates (NHPs). A major challenge is delivering intense, spatiotemporally-precise, patterned photostimulation across large volumes in deep tissue. Such stimulation is critical, for example, to modulate selectively deep-layer corticocortical feedback circuits. To address this need, we have developed the Utah Optrode Array (UOA), a 10×10 glass needle waveguide array fabricated atop a novel opaque optical interposer, and bonded to an electrically addressable µLED array. In vivo experiments with the UOA demonstrated large-scale, spatiotemporally precise, activation of deep circuits in NHP cortex. Specifically, the UOA permitted both focal (confined to single layers/columns), and widespread (multiple layers/columns) optogenetic activation of deep layer neurons, as assessed with multi-channel laminar electrode arrays, simply by varying the number of activated µLEDs and/or the irradiance. Thus, the UOA represents a powerful optoelectronic device for targeted manipulation of deep-layer circuits in NHP models.
more »
« less
- Award ID(s):
- 1755431
- PAR ID:
- 10495417
- Publisher / Repository:
- Nature Publishing Group
- Date Published:
- Journal Name:
- Communications Biology
- Volume:
- 7
- Issue:
- 1
- ISSN:
- 2399-3642
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
ABSTRACT Optogenetics has transformed the study of neural circuit function, but limitations in its application to species with large brains, such as non-human primates (NHPs), remain. A major challenge in NHP optogenetics is delivering light to sufficiently large volumes of deep neural tissue with high spatiotemporal precision, without simultaneously affecting superficial tissue. To overcome these limitations, we recently developed and testedin vivoin NHP cortex, the Utah Optrode Array (UOA). This is a 10×10 array of penetrating glass shanks, tiling a 4×4mm2area, bonded to interleaved needle-aligned and interstitial µLED arrays, which allows for independent photostimulation of deep and superficial brain tissue. Here, we investigate the acute biological response to UOA implantation in NHP cortex, with the goal of optimizing device design for reduced insertion trauma and subsequent chronic response. To this goal, we systematically vary UOA shank diameter, surface texture, tip geometry, and insertion pressure, and assess their effects on astrocytes, microglia, and neuronal viability, following acute implantation. We find that UOAs with shanks of smaller diameter, smooth surface texture and round tips cause the least damage. Higher insertion pressures have limited effects on the inflammatory response, but lead to greater tissue compression. Our results highlight the importance of balancing shank diameter, tip geometry, and insertion pressure in UOA design for preserving tissue integrity and improving long-term UOA performance and biocompatibility.more » « less
-
Abstract Deep neural networks such as GoogLeNet, ResNet, and BERT have achieved impressive performance in tasks such as image and text classification. To understand how such performance is achieved, we probe a trained deep neural network by studying neuron activations, i.e.combinations of neuron firings, at various layers of the network in response to a particular input. With a large number of inputs, we aim to obtain a global view of what neurons detect by studying their activations. In particular, we develop visualizations that show the shape of the activation space, the organizational principle behind neuron activations, and the relationships of these activations within a layer. Applying tools from topological data analysis, we presentTopoAct, a visual exploration system to study topological summaries of activation vectors. We present exploration scenarios usingTopoActthat provide valuable insights into learned representations of neural networks. We expectTopoActto give a topological perspective that enriches the current toolbox of neural network analysis, and to provide a basis for network architecture diagnosis and data anomaly detection.more » « less
-
Abstract Orographically‐locked diurnal convection involves interactions between local circulation and the thermodynamic environment of convection. Here, the relationships of convective updraft structures over orographic precipitation hotspots and their upstream environment in the TaiwanVVM large‐eddy simulations are analyzed for the occurrence of the orographic locking features. Strong convective updraft columns within heavily precipitating, organized systems exhibit a mass flux profile gradually increasing with height through a deep lower‐tropospheric inflow layer. Enhanced convective development is associated with higher upstream moist static energy (MSE) transport through this deep‐inflow layer via local circulation, augmenting the rain rate by 36% in precipitation hotspots. The simulations provide practical guidance for targeted observations within the most common deep‐inflow path. Preliminary field measurements support the presence of high MSE transport within the deep‐inflow layer when organized convection occurs at the hotspot. Orographically‐locked convection facilitate both modeling and field campaign design to examine the general properties of active deep convection.more » « less
-
We consider the verification of input-relational properties defined over deep neural networks (DNNs) such as robustness against universal adversarial perturbations, monotonicity, etc. Precise verification of these properties requires reasoning about multiple executions of the same DNN. We introduce a novel concept of difference tracking to compute the difference between the outputs of two executions of the same DNN at all layers. We design a new abstract domain, DiffPoly for efficient difference tracking that can scale large DNNs. DiffPoly is equipped with custom abstract transformers for common activation functions (ReLU, Tanh, Sigmoid, etc.) and affine layers and can create precise linear cross-execution constraints. We implement an input-relational verifier for DNNs called RaVeN which uses DiffPoly and linear program formulations to handle a wide range of input-relational properties. Our experimental results on challenging benchmarks show that by leveraging precise linear constraints defined over multiple executions of the DNN, RaVeN gains substantial precision over baselines on a wide range of datasets, networks, and input-relational properties.more » « less
