NeuroSpark

A matlab tool for efficient searchlight analysis

NeuroSpark is a versatile MATLAB toolbox designed for conducting fast searchlight analyses on neuroimaging data. It facilitates the extraction of spheres from arbitrary regions of interest the performance of searchlight analyses.

Features

• Efficient Searchlight Analysis: Conduct searchlight analyses on neuroimaging data quickly and efficiently.
• Customizable ROIs: Define and load predefined ROIs using flexible pattern matching.
• Modular Architecture: Easy integration of custom analysis methods and additional metadata.
• Flexible Options: Customize analysis parameters such as padding for smoothing regions and cortex.

TODO:

• init && working! 🚀
• add fancy images to readme
• add drawing nifti images functions to visualize regions and overall results

Installation

Clone the repository to your local machine:

git clone https://github.com/mlionello/NeuroSpark.git

Setup

Configure constants.m according to where your atlas, and derivatives are placed

Usage

Use the example neurospark.m function to perform searchlight analyses. Customize the analysis parameters and options as needed.

[results, center_indices, full_indices] = neurospark(roi1_pattern, roi2_pattern, analyser, mdata, opts);

[results, center_indices, full_indices] = neurospark(["V1", "V4"], ["premtoro"], @perform_lr_on_searchlight, 'searchlight_radius', 3, 'searchlight_padding', 1, 'targetPropInVx', 0.7);

[results, center_indices, full_indices] = neurospark(["attention"], "allbrain", @perform_svm_on_searchlight, struct('reduction_method', @max));

Documentation

Refer to the function headers and comments within the codebase for detailed usage instructions.

Contributions

Contributions are welcome! If you find any bugs or have suggestions for improvement, please open an issue or submit a pull request.