Skip to content

Document the historical basis for TE = 30 ms at 3T - #104

Merged
tsalo merged 3 commits into
mainfrom
te-convention-history
Nov 17, 2026
Merged

tsalo merged 3 commits into
mainfrom
te-convention-history

Conversation

@tsalo

@tsalo tsalo commented Nov 17, 2026

Copy link
Copy Markdown
Member

Closes #92.

What changed

  • New section in content/fMRI_Sequences.md: "Why single-echo fMRI uses TE ≈ 30 ms at 3T"
    • Derives why BOLD sensitivity peaks at TE = T2*, citing experimental confirmation (Bandettini et al., 1994; Menon et al., 1993; van der Zwaag et al., 2009).
    • Reports measured gray matter T2* at 3T (about 40-55 ms; Wansapura et al., 1999; Robinson et al., 2004). Adds a table of typical cortical T2* and common single-echo TEs at 1.5T, 3T and 7T, with example protocols (HCP, ABCD, UK Biobank, 7T HCP).
    • Explains that the conventional TE is shorter than cortical T2* because of signal dropout near air-tissue boundaries (Ojemann et al., 1997; Gorno-Tempini et al., 2002; Deichmann et al., 2002; Stöcker et al., 2006). The flat sensitivity curve means the cost elsewhere is small; physiological noise flattens it further (Krüger & Glover, 2001; Triantafyllou et al., 2005; Fera et al., 2004).
    • Adds a figure of relative BOLD sensitivity versus TE for a typical cortical voxel (T2* = 50 ms) and a dropout voxel (T2* = 20 ms).
    • Closes with what this means for multi-echo acquisition and optimal combination.
  • Replaced a claim in the same chapter that 3T BOLD contrast is "largest at approximately 28ms" with a pointer to the new section.
  • Corrected content/01_Optimal_Combination_with_t2smap.md, which said gray matter T2* at 3T is "roughly 30-40 ms".
  • Added 15 references to content/references.bib.

Notes for reviewers

  • I checked the papers' abstracts on PubMed, and the DOIs of the four protocol papers against Crossref. The protocol TEs (HCP 33.1 ms, ABCD 30 ms, UK Biobank 39 ms, 7T HCP 22.2 ms) and the typical TE ranges per field strength are from memory, so please check them.
  • The van der Zwaag et al. (2009) T2* values (about 80, 55 and 33 ms) are derived: ΔR2* divided by the reported ΔR2*/R2* ratios. The paper does not state them directly.
  • fMRI_Sequences.md has a new code cell, so after merging the execution cache needs republishing (make publish-cache) before main deploys. The rest of the book was already cached, and the local build succeeded with no warnings.

🤖 Generated with Claude Code

Add a section to the fMRI sequences chapter explaining that the
conventional single-echo TE (about 30 ms at 3T, 40-50 ms at 1.5T,
20-25 ms at 7T) is shorter than the gray matter T2* at each field
strength, and that the shorter TE trades a small loss of BOLD
sensitivity in typical cortex for much less dropout near air-tissue
boundaries. Include a BOLD sensitivity figure, the supporting
literature, and correct the optimal combination chapter's claim that
gray matter T2* is 30-40 ms at 3T.

Closes #92.
Frame the dropout explanation for the 30 ms convention as an inference
from the evidence, since no cited source documents how the convention
arose. Replace the 55-75% TE-to-T2* range, which the table did not
support at every field strength, with a looser description. Add the
missing suffix to William S. Ball, Jr. in the Wansapura et al. entry.
Claude Code session worktrees live under .claude/worktrees/, and Sphinx
otherwise tries to build their files, which breaks local builds.
@tsalo
tsalo merged commit b3ee8f4 into main Nov 17, 2026
2 checks passed
@tsalo
tsalo deleted the te-convention-history branch November 17, 2026 20:26
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Document historical basis for TE=30 ms (in 3T) convention

1 participant