[Eeglablist] Hierarchical Event Descriptor (HED) system - request for endorsements

Scott Makeig smakeig at gmail.com
Tue Aug 17 15:31:53 PDT 2021

Annotating more precisely the nature of events in neurobehavioral and other
time series data is an essential step in advancing understanding of brain
dynamics, particularly to allow cross-study analyses using new AI methods.
The Hierarchical Event Descriptor (HED) system, first presented by Nima
Bigdely-Shamlo at SCCN in 2012, is the only system we are aware of that
provides a practical path for human-readable *and* machine-actionable
annotation of events. For this reason, in 2019 the BIDS Steering Committee
included HED in the top-level BIDS data structure, meaning it is now
specifically allowed in BIDS data of all modalities.

Kay Robbins and myself, with Dung Truong, and Arno Delorme have been
pushing development of the HED system development, most recently publishing
the specification document and base schema vocabulary for third-generation
HED (explorable at *hedtags.org/display_hed.html
<https://urldefense.proofpoint.com/v2/url?u=http-3A__hedtags.org_display-5Fhed.html&d=DwIFaQ&c=-35OiAkTchMrZOngvJPOeA&r=kB5f6DjXkuOQpM1bq5OFA9kKiQyNm1p6x6e36h3EglE&m=3G_D14_LdvZ1MoRMpSEmghkPOli3pC7a6VVAlnHAKGs&s=xSIpRMkXEBwyaEUd-7OX6RzT5tvHHqLZxUHr9IfZhC0&e= >*), and are now soliciting letters of
support for our submission of an NIH grant proposal, *Hierarchical Event
Descriptors (HED): a system for machine-actionable annotation of events and
experimental structure in neurobehavioral data.*  We have also continued to
build the HED tool infrastructure needed to support HED annotation and
validation, including implementing a HED validator function in the BIDS
validation software. Our next tasks are to tag more datasets and to write
tutorials on use of HED.

Concrete case studies have demonstrated how HED can document not only what
happened (in detail) during the data recording, but also the experimental
design and control variable context in which the event occurred, all in a
compact, human readable and machine actionable format. In addition, during
the last year we have continued to focus on making human use of HED and HED
annotation as simple and natural as possible, as we realize this is key to
wider adoption.

      One new feature we have added to HED is the concept (and syntax) for
extending the basic HED annotation vocabulary through addition of *HED
Library Schemas* to the base HED schema (like adding libraries in python).
Library schemas are intended to feature terminology in common use within a
particular research field or subfield. We are happy to say that three
groups are now building or planning HED library schema: (1) for language,
(2) for cinematic stimulation, and (3) to provide an open-source system for
recording and analyzing stored and shared data from clinical EEG
examinations, by adapting the internationally recognized SCORE vocabulary
used by clinical neurophysiologists to describe, e.g., alpha bursts, ictal
events, etc. This SCORE effort is led by Dora Hermes Miller at The Mayo
Clinics, Rochester MN.  Another group in Salzberg plans to use HED in
building an Austrian fMRI data resource.

      But the work of developing, maintaining, and promoting use of HED
annotation for stored and shared neuroimaging time series data is just
beginning – and important technical programming and support tasks need
funding – hence the importance of our proposed project to continue

The NIH proposal deadline is *Sept. 2,* so we will need to put together
letters of support for the proposal by August 30, latest. Showing that
there is indeed support for building and promulgating HED in the
neuroimaging research community is, we believe, essential for our proposal
to be seen by the review panel for the important effort we believe it to

For more information about HED (3G), see our latest ms. at


To send a letter expressing support for the HED project, expand on the text
below and send by email to smakeig at ucsd.edu.

Dear Drs. Makeig and Robbins,

I am writing in support of your proposed *Hierarchical Event Descriptors
(HED): a system for machine-actionable annotation of events and
experimental structure in neurobehavioral data* in response to NIH
RFA-MH-20-128 (*BRAIN Initiative: Standards to Define Experiments Related
to the BRAIN Initiative*). ...

Thanking you for your consideration,        Scott
Scott Makeig, Research Scientist and Director, Swartz Center for
Computational Neuroscience, Institute for Neural Computation, University of
California San Diego, La Jolla CA 92093-0559, http://sccn.ucsd.edu/~scott

More information about the eeglablist mailing list