SignalPlant: an open signal processing software platform

Area of Science:

• Biomedical Engineering
• Signal Processing
• Neuroscience

Background:

• Modern acquisition systems generate large biosignal datasets (e.g., whole-day EEG recordings).
• Visual inspection and labeling of these large datasets are hindered by slow rendering times in current software.
• Efficient tools are needed for interactive analysis of extensive multichannel physiological data.

Purpose of the Study:

• To develop SignalPlant, a standalone application for efficient signal inspection, labeling, and processing.
• To provide researchers with a tool for fast, interactive work with large multichannel biosignal records.
• To overcome the latency issues associated with visualizing extensive physiological data.

Main Methods:

• Development of SignalPlant as a standalone, plugin-extensible application.
• Implementation of optimized rendering algorithms for large signal datasets.
• Comparative analysis of rendering latency against established software (e.g., EEGLAB).

Main Results:

• SignalPlant demonstrates significantly faster rendering speeds for large sample sizes compared to EEGLAB (e.g., 163x faster for 75 million samples).
• The software is free, requires no external computation software dependencies, and is adaptable via plugins.
• Facilitates faster and more interactive analysis of electroencephalograph (EEG), electrocardiograph (ECG), and similar biosignals.

Conclusions:

• SignalPlant offers a substantial improvement in processing speed for large biosignal datasets.
• Its standalone nature, speed, and extensibility make it a valuable tool for researchers in neuroscience and biomedical engineering.
• The application enhances the efficiency of visual inspection and labeling tasks for multichannel physiological recordings.