A major theme of the book is that you cannot analyze what you cannot see. It emphasizes the importance of inspecting your data at every step—before filtering, after filtering, after epoching—ensuring you don't automate the production of garbage results.
Dr. Mike X. Cohen is an Assistant Professor in the Donders Institute for Brain, Cognition, and Behavior at Radboud University and University Medical Center in Nijmegen, the Netherlands. He leads the Synchronization in the Neural Systems research group, and his work focuses on using state-of-the-art neuroscience methods to understand the mechanisms and implications of brain circuit dynamics.
An expert-level guide on focuses on how researchers extract meaningful signals from brain activity recordings. This field sits at the intersection of neuroscience, signal processing, and data science, transforming raw voltage fluctuations into insights about human cognition.
: Run Independent Component Analysis (ICA) to isolate and remove eye blinks, muscle tension, and cardiac artifacts. A major theme of the book is that
Standard t-tests assume independent data points. Neural data is autocorrelated (tomorrow’s brain state is similar to today’s). The book introduces non-parametric permutation testing and cluster-based correction for multiple comparisons (via the FieldTrip toolbox).
Don't just download the PDF to let it sit on your hard drive. Work through the examples. Write the code. Plot the figures. As Cohen writes in the preface: “The goal is not to get through the book. The goal is to get the book through you.”
You do not need to write these complex mathematical algorithms from scratch. The neuroscience community has built robust, open-source toolboxes: Mike X
To extract features from neural time series data, researchers rely heavily on three mathematical pillars. Understanding the mechanics behind these transforms is crucial before writing any analysis scripts. A. The Fourier Transform
Applies Fourier transforms to sliding time windows.
The book was originally written alongside a robust library of raw MATLAB scripts. It teaches you to build your own analysis tools from scratch rather than relying blindly on black-box graphical user interface (GUI) toolboxes. An expert-level guide on focuses on how researchers
Neural time series data analysis has become an essential tool in understanding the complex dynamics of brain function and behavior. With the increasing availability of large-scale neural data, there is a growing need for robust and reliable methods to analyze and interpret these data. In this article, we will provide an in-depth overview of the theory and practice of analyzing neural time series data, with a focus on the latest advances and techniques in the field.
: Principal Components Analysis (PCA), surface Laplacian spatial filters, and cross-frequency coupling.
His research publications span topics including generalized eigendecomposition for denoising, contrast enhancement, and dimension reduction in multichannel electrophysiology, as well as midfrontal theta phase coordination. He has also published another MIT Press title, .
: What are the underlying rhythms (e.g., Alpha, Beta, Gamma)?