Skip to content
Resource Development Background

Resource Development

To lower the barrier to entry into systems neuroscience, I have been developing both recording devices and recorded data analysis tutorials. I believe that such open knowledge will broaden the foundation of systems neuroscience research and lead to innovative results.

RefFlow for Codex

Codex Skill

RefFlow is a Codex skill built to start from a source repository and a literature library instead of a generic prompt. It works from PDFs related to a specific project or research question, then generates targeted paper summaries, reviews, implementation proposals, and executable code in one continuous workflow.

Automated Spike Sorting Pipeline

Spike sorting is a notoriously time-consuming bottleneck in in vivo electrophysiology. This pipeline performs automated waveform curation to eliminate noise and artifacts from putative clusters, enabling robust and efficient spike sorting.

pyNeuroscope

A Python-based tool for exploring neuroscience recordings and building lightweight inspection workflows around neural data. It is designed as a practical environment for looking through signals, annotations, and analysis outputs in a more flexible way.

Sleepy

Software for online sleep-state classification, designed to estimate sleep stages from ongoing recordings in real time.

Microdrive for Multi-Site Tetrode Recording

Despite the advent of high-density silicon probes, simultaneously recording from multiple, distinct subcortical regions in freely behaving mice remains challenging due to geometric and weight constraints. This customizable microdrive overcomes these limits and includes a complete fabrication tutorial.

Neural Dynamics Analysis Tutorials

Active Development

Designed for students and researchers newly acquiring in vivo neural recordings, these comprehensive tutorials serve as an accessible, hands-on introduction to essential coding practices and fundamental methodologies for analyzing neural dynamics.

Low Computational-Cost Cell Detection (LCCD)

When capturing massive fields of view (e.g., 3 × 3 mm), robust and efficient cell detection is essential, as without it, the full potential of the microscope cannot be realized. Optimizing algorithms from Ito et al. (Neurosci Res., 2022), this method enhances processing speed for motion correction and soma detection in large-scale two-photon calcium imaging.

Section Isolator

Whole-slide microscopes often capture multiple tissue sections simultaneously within a single panoramic image. This tool automatically isolates and crops each individual brain slice, streamlining subsequent atlas registration and downstream spatial analysis.