Biology subjects

Soldado-Magraner, S.

Publications and source records attributed to Soldado-Magraner, S..

3 recordsLinked to original sources

bioRxiv2026

STIMscope: centimeter-scale all-optical imaging and patterned optogenetic manipulation at single-cell resolution

Linking observation to intervention at cellular resolution makes it possible to move from measuring network activity to testing the contribution of defined neurons or ensembles within the same preparation. All-optical probing provides this capability by combining fluorescence-based readout with targeted optogenetic manipulation. Yet the platforms that deliver this capability remain complex, expensive, and difficult to maintain, requiring specialized expertise that has confined them to a small number of laboratories. They also typically provide fields of view too limited for studies of large, distributed neuronal populations. We address these constraints with the Spatiotemporal Illumination Microscope (STIMscope), a one-photon benchtop platform that integrates large-aperture tandem optics with a small-pixel back-illuminated CMOS sensor, a digital micromirror device for patterned illumination, and a GPU-based processing unit coordinated by a microcontroller for hardware-level synchronization. Ray-tracing simulations and point-spread-function measurements confirm cellular-scale resolution, with imaging lateral FWHM of 5.6 {micro}m at the field center and 5.8 {micro}m at the edge, and excitation lateral FWHM of 5.8 {micro}m at the center and 6.2 {micro}m at the edge, supporting fields of view as large as 14 mm x 11 mm in the demagnified configuration. The accompanying Closed-loop ready Real-time Imaging and Stimulation Pipeline (CRISPI) provides GPU-accelerated calibrated mask projection (26.3 ms latency), online ROI trace extraction, and modular ZeroMQ-based control, with a measured imaging-to-stimulation loop benchmark of 91.6 ms. We validate STIMscope in fixed mouse brain tissue, live iPSC-derived human neuronal cultures, and ex vivo organotypic slices of mouse auditory cortex. In organotypic slices, we show that both static and spatiotemporal stimulus identity can be decoded from population activity, revealing reservoir-like population dynamics, and that this decodability remains stable in the same neuronal population over hours. We further show that post-stimulus activity retains information about recent stimuli for several seconds, consistent with short-term memory dynamics. With a bill of materials under $5,000 USD and all mechanical designs, firmware, and software released open-source, STIMscope makes all-optical neuroscience experiments a routine capability accessible to laboratories without specialized optical engineering expertise.

bioengineering↗

bioRxiv2023

Cross-homeostatic plasticity enables analog neuromorphic circuits to exhibit robust computational primitives

AbstractMany neural computations emerge from self-sustained patterns of activity in recurrent neural circuits, which rely on balanced excitation and inhibition. Neuromorphic electronic circuits that use the physics of silicon to emulate neuronal dynamics represent a promising approach for implementing the brains computational primitives, including self-sustained neural activity. However, achieving the same robustness of biological networks in neuromorphic computing systems remains a challenge, due to the high degree of heterogeneity and variability of their analog components. Inspired by the strategies used by real cortical networks, we apply a biologically-plausible cross-homeostatic learning rule to balance excitation and inhibition in neuromorphic implementations of spiking recurrent neural networks. We demonstrate how this learning rule allows the neuromorphic system to work in the presence of device mismatch and to autonomously tune the spiking network to produce robust, self-sustained, fixed-point attractor dynamics with irregular spiking in an inhibition-stabilized regime. We show that this rule can implement multiple, coexisting stable memories, with emergent soft-winner-take-all (sWTA) dynamics, and reproduce the so-called "paradoxical effect" widely observed in cortical circuits. In addition to validating neuroscience models on a substrate that shares many similar properties and limitations with biological systems, this work enables the construction of ultra-low power, mixed-signal neuromorphic technologies that can be automatically configured to compute reliably, despite the large on-chip and chip-to-chip variability of their analog components.

neuroscience↗

bioRxiv2021

Orchestrated Excitatory and Inhibitory Learning Rules Lead to the Unsupervised Emergence of Up-states and Balanced Network Dynamics

Self-sustaining neural activity maintained through local recurrent connections is of fundamental importance to cortical function. We show that Up-states--an example of self-sustained, inhibition-stabilized network dynamics--emerge in cortical circuits across three weeks of ex vivo development, establishing the presence of unsupervised learning rules capable of generating self-sustained dynamics. Previous computational models have established that four sets of weights (WE[<-]E, WE[<-]I, WI[<-]E, WI[<-]I) must interact in an orchestrated manner to produce Up-states, but have not addressed how a family of learning rules can operate in parallel at all four weight classes to generate self-sustained inhibition-stabilized dynamics. Using numerical and analytical methods we show that, in part due to the paradoxical effect, standard homeostatic rules are only stable in a narrow parameter regime. In contrast, we show that a family of biologically plausible learning rules based on "cross-homeostatic" plasticity robustly lead to the emergence of self-sustained, inhibition-stabilized dynamics.

neuroscience↗