Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Longer scans boost prediction and cut costs in brain-wide association studies

Ooi, Leon Qi Rong ; Orban, Csaba ; Zhang, Shaoshi ; Nichols, Thomas E. ; Tan, Trevor Wei Kiat ; Kong, Ru ; Marek, Scott ; Dosenbach, Nico U.F. ; Laumann, Timothy O. and Gordon, Evan M. , et al. (2025) In Nature 644(8077). p.731-740
Abstract

A pervasive dilemma in brain-wide association studies1 (BWAS) is whether to prioritize functional magnetic resonance imaging (fMRI) scan time or sample size. We derive a theoretical model showing that individual-level phenotypic prediction accuracy increases with sample size and total scan duration (sample size × scan time per participant). The model explains empirical prediction accuracies well across 76 phenotypes from nine resting-fMRI and task-fMRI datasets (R2 = 0.89), spanning diverse scanners, acquisitions, racial groups, disorders and ages. For scans of ≤20 min, accuracy increases linearly with the logarithm of the total scan duration, suggesting that sample size and scan time are initially interchangeable.... (More)

A pervasive dilemma in brain-wide association studies1 (BWAS) is whether to prioritize functional magnetic resonance imaging (fMRI) scan time or sample size. We derive a theoretical model showing that individual-level phenotypic prediction accuracy increases with sample size and total scan duration (sample size × scan time per participant). The model explains empirical prediction accuracies well across 76 phenotypes from nine resting-fMRI and task-fMRI datasets (R2 = 0.89), spanning diverse scanners, acquisitions, racial groups, disorders and ages. For scans of ≤20 min, accuracy increases linearly with the logarithm of the total scan duration, suggesting that sample size and scan time are initially interchangeable. However, sample size is ultimately more important. Nevertheless, when accounting for the overhead costs of each participant (such as recruitment), longer scans can be substantially cheaper than larger sample size for improving prediction performance. To achieve high prediction performance, 10 min scans are cost inefficient. In most scenarios, the optimal scan time is at least 20 min. On average, 30 min scans are the most cost-effective, yielding 22% savings over 10 min scans. Overshooting the optimal scan time is cheaper than undershooting it, so we recommend a scan time of at least 30 min. Compared with resting-state whole-brain BWAS, the most cost-effective scan time is shorter for task-fMRI and longer for subcortical-to-whole-brain BWAS. In contrast to standard power calculations, our results suggest that jointly optimizing sample size and scan time can boost prediction accuracy while cutting costs. Our empirical reference is available online for future study design (https://thomasyeolab.github.io/OptimalScanTimeCalculator/index.html).

(Less)
Please use this url to cite or link to this publication:
author
; ; ; ; ; ; ; ; and , et al. (More)
; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; and (Less)
author collaboration
organization
publishing date
type
Contribution to journal
publication status
published
subject
in
Nature
volume
644
issue
8077
pages
10 pages
publisher
Nature Publishing Group
external identifiers
  • pmid:40670782
  • scopus:105013871305
ISSN
0028-0836
DOI
10.1038/s41586-025-09250-1
language
English
LU publication?
yes
id
69f18da1-4234-4515-bbcc-a461eed43238
date added to LUP
2025-11-05 13:37:22
date last changed
2025-12-18 05:25:05
@article{69f18da1-4234-4515-bbcc-a461eed43238,
  abstract     = {{<p>A pervasive dilemma in brain-wide association studies<sup>1</sup> (BWAS) is whether to prioritize functional magnetic resonance imaging (fMRI) scan time or sample size. We derive a theoretical model showing that individual-level phenotypic prediction accuracy increases with sample size and total scan duration (sample size × scan time per participant). The model explains empirical prediction accuracies well across 76 phenotypes from nine resting-fMRI and task-fMRI datasets (R<sup>2</sup> = 0.89), spanning diverse scanners, acquisitions, racial groups, disorders and ages. For scans of ≤20 min, accuracy increases linearly with the logarithm of the total scan duration, suggesting that sample size and scan time are initially interchangeable. However, sample size is ultimately more important. Nevertheless, when accounting for the overhead costs of each participant (such as recruitment), longer scans can be substantially cheaper than larger sample size for improving prediction performance. To achieve high prediction performance, 10 min scans are cost inefficient. In most scenarios, the optimal scan time is at least 20 min. On average, 30 min scans are the most cost-effective, yielding 22% savings over 10 min scans. Overshooting the optimal scan time is cheaper than undershooting it, so we recommend a scan time of at least 30 min. Compared with resting-state whole-brain BWAS, the most cost-effective scan time is shorter for task-fMRI and longer for subcortical-to-whole-brain BWAS. In contrast to standard power calculations, our results suggest that jointly optimizing sample size and scan time can boost prediction accuracy while cutting costs. Our empirical reference is available online for future study design (https://thomasyeolab.github.io/OptimalScanTimeCalculator/index.html).</p>}},
  author       = {{Ooi, Leon Qi Rong and Orban, Csaba and Zhang, Shaoshi and Nichols, Thomas E. and Tan, Trevor Wei Kiat and Kong, Ru and Marek, Scott and Dosenbach, Nico U.F. and Laumann, Timothy O. and Gordon, Evan M. and Yap, Kwong Hsia and Ji, Fang and Chong, Joanna Su Xian and Chen, Christopher and An, Lijun and Franzmeier, Nicolai and Roemer-Cassiano, Sebastian N. and Hu, Qingyu and Ren, Jianxun and Liu, Hesheng and Chopra, Sidhant and Cocuzza, Carrisa V. and Baker, Justin T. and Zhou, Juan Helen and Bzdok, Danilo and Eickhoff, Simon B. and Holmes, Avram J. and Yeo, B. T.Thomas and Jack, Clifford R.}},
  issn         = {{0028-0836}},
  language     = {{eng}},
  number       = {{8077}},
  pages        = {{731--740}},
  publisher    = {{Nature Publishing Group}},
  series       = {{Nature}},
  title        = {{Longer scans boost prediction and cut costs in brain-wide association studies}},
  url          = {{http://dx.doi.org/10.1038/s41586-025-09250-1}},
  doi          = {{10.1038/s41586-025-09250-1}},
  volume       = {{644}},
  year         = {{2025}},
}