Feasibility and Validity of Ultra- Low-Field MRI for Measurement of Regional Infant Brain Volumes in Structures Associated with Antenatal Maternal Anemia

  1. Jessica E. Ringshaw
  2. Niall J. Bourke
  3. Michal R. Zieff
  4. Catherine J. Wedderburn
  5. Chiara Casella
  6. Layla E. Bradford
  7. Simone R. Williams
  8. Donna Herr
  9. Marlie Miles
  10. Jonathan O’Muircheartaigh
  11. Carly Bennallick
  12. Khula South Africa Data Collection Team
  13. Sean Deoni
  14. Dan J. Stein
  15. Daniel C. Alexander
  16. Derek K. Jones
  17. Steven C.R. Williams
  18. Kirsten A. Donald
Read the full article See related articles

Listed in

This article is not in any list yet, why not save it to one of your lists.
Log in to save this article

Abstract

Introduction

The availability of ultra- low-field (ULF) magnetic resonance imaging (MRI) has the potential to improve neuroimaging accessibility in low-resource settings. However, the utility of ULF MRI in detecting child brain changes associated with anemia is unknown.

Aim

The aim of this study was to assess the comparability of 3T high-field (HF) and 64mT ULF volumes in infants for brain regions associated with antenatal maternal anemia.

Method

This neuroimaging sub-study is nested within Khula South Africa, a population-based birth cohort. Pregnant women were enrolled antenatally and postnatally, and mother-child dyads ( n = 394) were followed prospectively at approximately 3, 6, 12 and 18 months. A sub-group of infants was scanned on 3T and 64mT MRI systems across study visits and images were segmented using MiniMORPH. Correlations and concordance coefficients were used to cross-validate HF and ULF infant brain volumes for the caudate nucleus, putamen, and corpus callosum.

Results

78 children (53.85% male) had paired HF ( Mean [ SD ] age = 9.64 [5.26] months) and ULF ( Mean [ SD ] age = 9.47 [5.32] months) datasets. Results indicated strong agreement between systems for intracranial volume (ICV; r = 0.96, ρ ccc = 0.95), and brain regions of interest in anemia including the caudate ( r = 0.89, ρ ccc = 0.86), putamen ( r = 0.97, ρ ccc = 0.96), and corpus callosum ( r = 0.87, ρ ccc = 0.79).

Conclusion

This cross-validation study demonstrates excellent correspondence between 3T and 64mT volumes for infant brain regions implicated in antenatal maternal anemia. Findings validate the use of ULF MRI for paediatric neuroimaging on anemia in Africa.

HIGHLIGHTS

  • This cross-validation study is the first to compare HF and ULF volume estimates for infant brain regions previously found to be associated with antenatal maternal anemia within the first two years of life.

  • Key findings of this research demonstrate linear associations and strong agreement between HF and ULF volume estimates for the caudate nucleus, putamen, and corpus callosum in infants between 3-18 months of age. Improved correspondence between HF and ULF MRI was observed in older infants, particularly for basal ganglia structures.

  • These novel findings validate the use of ULF MRI for paediatric neuroimaging work on antenatal maternal anemia and other prevalent health priorities in low- and middle-income countries.

GRAPHICAL ABSTRACT

This cross-validation study assessed the comparability of high-field (3T) and ultra- low-field (64mT) volumes in infants for brain regions associated with antenatal maternal anemia. Key findings demonstrated strong agreement between HF and ULF volume estimates for the caudate nucleus, putamen, and corpus callosum in infants between 3-18 months of age.

Article activity feed

  1. Version published to 10.1101/2025.06.10.25329262v1 on medRxiv