Your Minds AI Signal Filter DLL

This project contains samples for referencing your MindsAI.dll which requires a license key that can be purchased at https://www.minds-applied.com/

Version

v1.0.1

Calling Card

Before calling any signal processing functions, you must initialize the provided MindsAI license like:

InitializeMindsAILicense("YOUR-LICENSE-KEY");

• This only needs to be called once prior to filtering, and should not be called continuously to prevent API request overload errors.
• If the key is active, data should continue streaming.
• If data does not continue streaming or is not filtered, you can print license details via:

GetMindsAILicenseMessag(); 

• This will show you the expiration date and whether or not the key is valid.

After initializing, you can call the filter like:

ApplyMindsAIFilter(
    double* input,      // Flattened 3D array: [segments][timepoints][channels]
    int segments,       // Number of independent data segments (e.g., 1-4 for real-time or higher for batch trials)
    int timepoints,     // Number of time samples per segment (e.g., 500 for 1 sec @ 500Hz)
    int channels,       // Number of input channels (e.g., EEG electrodes)
    double lambda,      // Regularization coefficient (e.g., 1e-5; tune for filtering strength or use ML-calibrated value)
    double* output      // Flattened 3D array: [segments][channels][timepoints] with filtered data
);

The lambda hyperparameter controls how much the MindsAI Filter modifies the original signal and should be input on a logarithmic scale between 0 and 0.1. A lower lambda value like 1e-9 causes the Filter to make bolder adjustments for more complex transformations that can highlight structure across channels, such as for real-time filtering with shorter segments or fewer timepoints. A higher lambda value like 1e-5 works best with more data (such as 60-second segments) for still helpful, but more conservative adjustments.

Acknowledgments

Brainflow