Javier Gonzalez-Castillo

Javier Gonzalez-Castillo

Senior Associate Scientist, NIMH, NIH. Views expressed here are my own, and not those of my employer.

Sitemap

A list of all the pages found on the site. For you robots out there is an XML version available for digesting as well.

Pages

About me

Javier Gonzalez-Castillo is a Senior Associate Scientist at the NIH conducting research on fMRI methods, resting state functional connectivity, realtime fMRI, the neural correlates of internally driven cognition Read more

Research

Current cognitive neuroscience research projects by Javier Gonzalez-Castillo Read more

talks

What is the ultimate sensitivity of fMRI: does the whole brain activate?

Published:

Original description of our work looking at the true extent of fMRI activity when signal-to-noise ratio is sufficiently high and we also take into account transitory and negative BOLD responses. Read more

</article> </div>

Realtime fMRI/fMRI Neurofeedback

Published:

Introductory talk to realtime fMRI and fMRI-based neurofeedback. Here I describe what is realtime fMRI, the details of the realtime fMRI system that we have running at the NIH, and a series of studies that takes advantage of this unique technology. Read more

</article> </div>

Multi-echo EPI for resting state and activaton based fMRI

Published:

Here I give an overview on our efforts at the time in developing methods for multi-echo fMRI, and how you can use multi-echo fMRI to perform automatic denoising. Read more

</article> </div>

Uncovering hidden activations using model-free analysis

Published:

Here I describe our work looking at unconventional BOLD responses such as negative and transient responses that accompany block task designs. Read more

</article> </div>

Unconventional fMRI methodology: multi-echo fMRI, connectivity dynamics, and fMRI neurofeedback

Published:

Overview of alternative ways to acquire and analyze fMRI data other than typical task-based experiments. Note that this talk is in Spanish language. Read more

</article> </div>

Machine Learning and fMRI

Published:

Introductory talk to machine learning (ML) for neuroimagers. Here I introduce basic concepts such as supervised and unsupervised learning, regression, classification, gradient descent, overfitting and testing. I next discuss a few examples of how ML can be applied to functional neuroimaing data. Read more

</article> </div>

Quantitative deconvolution of neuronal-related BOLD events with multi-echo sparse free paradigm mapping

Published:

Here I describe our work on the design and evaluation of a multi-echo based formulation of the hemodynamic deconvolution algorithm called free paradigm mapping. Read more

</article> </div>

Multi-echo EPI for task-based and resting-state fMRI

Published:

Introductory talk to multi-echo fMRI. Here I discuss what is multi-echo fmri, basic operations you can do with multi-echo data (e.g., optimal combination), and more advanced approaches such as me-based denoising (ME-ICA) and multi-ech based deconvolution (ME-SPFM). Read more

</article> </div>

Periods of discernible cognition contribute to dynamic functional connectivity during rest

Published:

This conference talk describes our work looking at how to extract information about on-going cognition during rest using time varying functional connectivity and hemodynamic deconvolution. Read more

</article> </div>

Dynamic connectivity: is it real? is it useful? how do we extract information?

Published:

Introductory talk to dynamic functional connectivity. Here I discuss what is dynamic functional connectivity, how do we typically measure it. It then goes into what it is known today about how to interpret results. Read more

</article> </div>

Cognitive correlates of BOLD resting-state dynamic functional connectivity

Published:

Overview of recent work looking at the cognitive correlates of time varying functional connectivity, as well as some thoughts on future directions. Read more

</article> </div>

Individual variability in resting-state fMRI [VIDEO AVAILABLE]

Published:

Overview of state-of-the-art regarding our understanding of sources of individual differences in resting-state fMRI. This talk first describes the primary sources of variability, and then provide some thoughts on how to address those sources depending on whether a given study targets “state-level” or “trait-level” properties of the brain. [] Read more

</article> </div>

Now that I have Multi-Echo data, what else can I do with it?

Published:

Multi-edho data are best known for enabling the automatic identification and removal of artifactual non-BOLD components, thereby improving data quality. However, the richer information provided by ME acquisitions can also be leveraged to enhance the sensitivity of deconvolution algorithms, assess data quality via functional connectivity across echoes, and estimate the BOLD-likelihood of signals of interest, such as the global signal. In this talk, I will highlight these lesser-known applications of multi-echo data. Read more

</article> </div>

Role Of In-Scanner Conscious Experience In Resting-State

Published:

Resting-state fMRI (rsfMRI) is widely used to identify altered functional connectivity (FC) in clinical populations, yet studies rarely consider how conscious experience during scanning influences FC. In this talk, I present evidence from over 400 annotated resting-state scans showing that differences in conscious experience systematically shape FC and can themselves be predicted from FC. These findings highlight the importance of accounting for in-scanner experience when interpreting rsfMRI data and demonstrate the value of annotating resting-state datasets with experiential information. Read more

</article> </div>

Realtime fMRI and Neurofeedback

Published:

This talk introduces the unique features of real-time fMRI in comparison with conventional offline fMRI experimentation. It discusses the differences between real-time and offline preprocessing, reviews common applications of real-time fMRI, and highlights several real-time fMRI and neurofeedback studies that serve as exemplary models for conducting this type of research. Read more

</article> </div>