October 11, 2018
CiNet 1F Conference Room
Research Director & CEO
Host : Daniel Callan （CiNet PI）
In this talk I will present and demonstrate the NeuroPype Suite Academic Edition, a suite of desktop applications made available for free to the scientific community with the goal of accelerating and streamlining the creation, deployment, and sharing of pipelines for real-time or batch processing and decoding of unimodal and multi-modal sensor data (e.g. EEG, ExG, intracranial electrophysiology, actigraphy, eye tracking, audio and video signals, etc). The suite includes a version of NeuroPype, an extensible Python-based dataflow programming environment, containing 250+ modular data processing and visualization routines (“nodes”) that can be configured and linked together to create and run pipelines for signal processing and analysis, BCI and signal classification, neuroimaging, closed-loop feedback and control, and more. NeuroPype integrates seamlessly with the open-source Lab Streaming Layer (LSL) protocol for data acquisition, synchronization, and I/O, and works out of the box with LSL-compatible hardware devices. An extensible framework enables users to add their own custom processing nodes in Python, while a RESTful API allows external applications to interface with NeuroPype to create, run and configure pipelines in real-time. While NeuroPype can be operated entirely programmatically, the suite also includes Pipeline Designer, an open-source visual programming application based on Orange which provides an intuitive drag-and-drop GUI for pipeline design and configuration through NeuroPype’s API. I will illustrate both local deployment of NeuroPype pipelines, as well as deployment on the NeuroScale cloud platform for streaming access to/from mobile and other internet-connected devices and for scalable batch processing.