SHYBRID: A Graphical Tool for Generating Hybrid Ground-Truth Spiking Data for Evaluating Spike Sorting Performance.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Signal Processing

Background:

• Spike sorting is crucial for analyzing extracellular neural recordings, identifying individual neuron action potentials.
• Numerous spike sorting algorithms exist, but choosing the right one for specific recording conditions remains challenging.

Purpose of the Study:

• To develop an open-source graphical tool for generating hybrid ground-truth data.
• To assist users in selecting appropriate spike sorting algorithms based on recording settings (brain region, device).
• To facilitate algorithm parameter tuning and deepen understanding of spike sorting methods.

Main Methods:

• The tool utilizes a data-driven modeling approach to create hybrid ground-truth datasets.
• Spike times from single units are repositioned on the recording probe to simulate virtual units with known spike times.
• The implementation is in Python, offering a user-friendly graphical interface.

Main Results:

• Enables efficient generation of hybrid ground-truth datasets.
• Provides a framework for informed decision-making when choosing between different spike sorting algorithms.
• Facilitates algorithm parameter optimization tailored to specific recording configurations.

Conclusions:

• The developed tool empowers neuroscientists to more effectively evaluate and utilize spike sorting algorithms.
• Improved spike sorting accuracy can be achieved by using tailored ground-truth data for algorithm selection and tuning.
• This resource enhances the understanding and application of computational methods in neural data analysis.