📄 hiv aids

Ultra-low-field MRI as a tool for measuring brain development in at-risk children in LMICS: feasibility, validity and clinical relevance.

This study demonstrates that ultra-low-field (64mT) MRI is a feasible and valid, scalable alternative to high-field (3T) MRI for measuring brain volumes and assessing early language development in HIV-exposed and unexposed children in low- and middle-income countries.

Original authors: Bradford, L. E., Ringshaw, J. E., Malaba, T. R., Bourke, N. J., Wedderburn, C. J., Williams, S. C., Deoni, S., Reynolds, H., Read, J., Read, L., Waitt, C., Mrubata, M., Stemmet, L.-A., Davel, L., Colb

Published 2026-06-05

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Bradford, L. E., Ringshaw, J. E., Malaba, T. R., Bourke, N. J., Wedderburn, C. J., Williams, S. C., Deoni, S., Reynolds, H., Read, J., Read, L., Waitt, C., Mrubata, M., Stemmet, L.-A., Davel, L., Colbers, A., Wang, D., Khoo, S., Myer, L., Donald, K. A.

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ⚕️ This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer

The Big Picture: A New, Affordable Camera for the Brain

Imagine trying to study how a child's brain grows, but you only have access to expensive, giant, high-tech cameras that cost millions of dollars and need a massive building to house them. This is the current reality for many low- and middle-income countries (LMICs).

This study tests a new, smaller, and much cheaper "camera" called Ultra-Low-Field (ULF) MRI. Think of it like comparing a professional, high-end DSLR camera (the standard 3T MRI) to a very high-quality smartphone camera (the 64mT ULF MRI). The researchers wanted to know: Can this cheaper, portable "smartphone camera" take pictures of a child's brain that are good enough to see the same details as the giant professional camera?

They also wanted to see if the pictures taken by this new camera could actually tell us something useful about how well a child is developing, specifically their ability to speak and understand language.

The "Test Subjects": A Vulnerable Group

The study focused on children in South Africa, specifically a group called CHEU (Children who are HIV-Exposed but Uninfected).

The Analogy: Imagine these children were born to mothers who took medicine to prevent passing a virus to them. The children didn't get the virus, but because they were exposed to the virus and the medicine while in the womb, they are at a slightly higher risk for developmental delays, particularly in language and movement.
The researchers compared these children to a group of children who were never exposed to HIV at all (CHU), to see if the new MRI could spot differences in brain structure between the two groups.

How They Did It: The "Sleeping Beauty" Protocol

Scanning a toddler's brain is hard because they can't sit still. Usually, doctors have to put children to sleep using strong sedation drugs, which is risky and expensive.

The Innovation: The team used a special technique to scan the children while they were in natural, non-sedated sleep.
The Analogy: Instead of forcing a child to sit still like a statue, they waited until the child was naturally napping, then gently placed them in the scanner. This made the process safer and more feasible for places without advanced medical anesthesia teams.

The Results: Do the Pictures Match?

The researchers took pictures of the same 45 children using both the giant "professional" camera (3T MRI) and the new "portable" camera (ULF MRI).

The Big Structures Matched Well:
- Analogy: If you look at a map of a country, both cameras could clearly see the big continents and large oceans.
- The Finding: The ULF MRI was excellent at measuring the brain's overall size and large areas like the "grey matter" (the brain's thinking tissue) and "white matter" (the wiring). It also did a great job seeing big sub-cities like the thalamus and caudate. The numbers from the cheap camera lined up very closely with the expensive one.
The Small Details Were a Bit Blurrier:
- Analogy: If you tried to read the name of a tiny street on a small map, it might look fuzzy.
- The Finding: For very small, complex structures deep in the brain (like the amygdala or hippocampus), the ULF MRI wasn't as sharp. It sometimes underestimated the size of these tiny areas. However, the correlation was still there, just not perfect.
The "Volume" Quirk:
- The Finding: The ULF camera tended to slightly "shrink" the grey matter in its measurements but slightly "inflate" the white matter. The researchers noted this is likely because the lower-resolution images make it harder to draw the exact line between grey and white tissue, similar to how a low-resolution photo might blur the edge of a shadow.

The Real-World Connection: Brain Size and Talking

The most exciting part of the study wasn't just comparing the cameras, but seeing if the pictures mattered for the children's lives.

The Finding: The brain volumes measured by the cheap ULF camera were strongly linked to how well the children could understand and use language.
The Analogy: It's like finding that the size of a specific engine part in a car predicts how fast the car can drive. The study found that children with larger volumes in certain brain areas (like the thalamus and hippocampus) tended to have better scores on language tests.
Why it matters: Since the ULF MRI could spot these connections just as well as the expensive MRI, it proves that this cheaper tool can actually help doctors and researchers understand which children might be at risk for language delays.

The Bottom Line

The paper concludes that this new, portable, low-cost MRI machine is a feasible and valid tool.

It works in resource-limited settings (like South Africa).
It doesn't require dangerous sedation drugs.
It produces brain measurements that are very similar to the gold-standard expensive machines.
It can successfully link brain structure to language development.

In short, the researchers successfully proved that you don't need a multi-million-dollar machine in a massive hospital to get useful, life-changing data about a child's brain development. You can do it with a portable, affordable device that fits in a standard room.

Technical Summary: Ultra-low-field MRI for Measuring Brain Development in LMICs

Problem Statement
Children in low- and middle-income countries (LMICs), particularly those who are HIV-exposed but uninfected (CHEU), face elevated risks of neurodevelopmental delays, including language and motor deficits. While high-field (HF) MRI (≥1.5T) is the gold standard for investigating the neurobiological basis of these delays, its application in LMICs is severely limited by high costs, infrastructure requirements, and operational complexity. Ultra-low-field (ULF) MRI (<100mT) has emerged as a potential affordable and portable alternative, yet its validity for measuring brain structure in early childhood and its ability to correlate with developmental outcomes in vulnerable populations remain underexplored.

Methodology
This study utilized a cross-sectional sub-study design nested within the DolPHIN-2 PLUS observational follow-up study conducted in Cape Town, South Africa.

Participants: The cohort included 45 children (mean age 45.6 months) with paired, usable T2-weighted MRI scans. The group consisted of 9 CHEU and 36 children who are HIV-unexposed (CHU).
Imaging Protocol:
- ULF MRI: Acquired using a Hyperfine Swoop 64mT system with an 8-channel head coil. T2-weighted images were acquired in three orthogonal planes and reconstructed into a single high-resolution isotropic image (1.5 mm³) using multi-resolution registration.
- HF MRI: Acquired using a Siemens Skyra 3T scanner with a 32-channel head coil, utilizing high-resolution T2-weighted sequences (1 mm³).
- Conditions: All scans were performed during natural, non-sedated sleep to ensure safety and feasibility in a resource-constrained setting.
Processing: Volumetric segmentation was performed using FreeSurfer v7.4.1 and SynthSeg on the Flywheel platform, adhering to the Desikan-Killiany atlas for global and subcortical regions of interest (ROI).
Developmental Assessment: Child development was measured using the Bayley Scales of Infant and Toddler Development, Third Edition (BSID-III), assessing cognitive, language (receptive/expressive), and motor domains.
Statistical Analysis: Agreement between modalities was assessed using Pearson's correlation coefficients ( $r$ ) and Lin's concordance correlation coefficients ( $P_{ccc}$ ). Associations between ULF-derived volumes and developmental outcomes were evaluated using partial correlations adjusted for sex and age.

Key Contributions

Feasibility Validation: The study demonstrates the successful implementation of paired ULF and HF MRI in preschool-aged children in an LMIC setting using non-sedated sleep protocols.
Volumetric Concordance: It provides empirical evidence comparing brain volume estimates derived from 64mT ULF MRI against 3T HF MRI across global and subcortical regions.
Clinical Relevance: It establishes the first link between ULF-derived brain volumes and developmental outcomes (specifically language) in a vulnerable pediatric population, validating the tool's utility beyond mere structural imaging.

Results

Imaging Agreement:
- Global Regions: Strong to very strong linear correlations were observed between ULF and HF volumes for total intracranial volume, total grey matter, and total white matter ( $r > 0.92$ ).
- Subcortical Structures: Strong correlations were found in larger structures such as the thalamus ( $r = 0.889$ ), caudate ( $r = 0.881$ ), and putamen ( $r = 0.860$ ). Moderate correlations were observed in the pallidum and hippocampus ( $r \approx 0.63$ ), while the amygdala showed weak correlation ( $r = 0.417$ ).
- Systematic Bias: ULF MRI systematically underestimated most grey matter volumes relative to HF MRI (differences ranging from -0.17% to -15.76%) but overestimated total white matter volume (+6.15%).
- Concordance: Only intracranial volume reached the threshold for substantial agreement ( $P_{ccc} > 0.95$ ). Most other regions showed fair to poor agreement, attributed to resolution limitations and partial volume effects in smaller structures.
Developmental Associations:
- ULF-derived global grey matter and white matter volumes showed significant positive associations with receptive and expressive communication scores.
- The strongest and most consistent associations were observed between subcortical grey matter volumes (including the thalamus, caudate, putamen, pallidum, hippocampus, and amygdala) and language domains. For example, hippocampal volume correlated with expressive communication ( $r = 0.59, p < 0.001$ ).
- These associations were confirmed using HF-derived volumes, reinforcing the biological validity of the ULF findings.

Significance and Claims
The paper claims that ULF MRI is a feasible and scalable tool for studying neurodevelopment in vulnerable pediatric populations within LMICs. While acknowledging that ULF MRI currently has lower spatial resolution and systematic volumetric biases compared to HF MRI, the authors assert that it produces valid volumetric estimates that correlate strongly with HF data for global and larger subcortical regions. Crucially, the study demonstrates that ULF-derived metrics capture biologically meaningful associations with early language development.

The authors position this work as a critical step toward addressing the global disparity in neuroimaging access. They suggest that ULF MRI can be utilized to investigate neurodevelopmental trajectories in at-risk groups, such as CHEU, and to monitor the long-term neurological implications of evolving antiretroviral therapy (ART) regimens. The study concludes that despite current limitations in resolution, the portability and cost-effectiveness of ULF MRI make it a viable alternative for expanding access to neuroimaging in resource-constrained settings, provided that future work continues to refine acquisition and processing pipelines.