Ian Kinsella

maskNMF: A denoise-sparsen-detect approach for extracting neural signals from dense imaging data

maskNMF improves demixing of dense calcium imaging data by denoising, sparsening, and using constrained non-negative matrix factorization to accurately recover neural activity.

Neuroscience Cloud Analysis As a Service

This paper introduces a cloud-based platform that provides automated, reproducible, and accessible data analysis tools for neuroscience, improving efficiency and reducing infrastructure barriers.

High fidelity estimates of spikes and subthreshold waveforms from 1-photon voltage imaging in vivo

This paper introduces a method to accurately estimate spikes and subthreshold waveforms from 1-photon voltage imaging in vivo, even with noise.

Localized semi-nonnegative matrix factorization (LocaNMF) of widefield calcium imaging data

This paper introduces LocaNMF, a method to decompose widefield calcium imaging data into interpretable spatial and temporal components for analyzing neural activity across mice.

EASE: EM-Assisted Source Extraction from calcium imaging data

The paper introduces EASE, a method that matches calcium imaging signals to EM-derived cell segments in mouse visual cortex, enabling dendritic functional analysis and denoising.

YASS: Yet Another Spike Sorter applied to large-scale multi-electrode array recordings in primate retina

YASS sorts spikes from multi-electrode arrays with > 500 electrodes using neural network detection and denoising, parallelized nonparametric Bayesian clustering, and matching-pursuit deconvolution, outperforming manual sorting on primate retina recordings.

All-optical electrophysiology reveals brain-state dependent changes in hippocampal subthreshold dynamics and excitability

The paper develops all-optical electrophysiology tools to record subthreshold dynamics and excitability in mouse hippocampus during behavior, revealing brain-state dependent changes in neuronal activity.

Penalized matrix decomposition for denoising, compression, and improved demixing of functional imaging data

~2-4x SNR gains and 20-300x compression on calcium and voltage imaging, from a patch-wise penalized matrix decomposition that splits signal from temporally uncorrelated noise without hyperparameter tuning.