Motion robust MR fingerprinting scan to image neonates with prenatal opioid exposure

Dan Ma, Chaitra Badve, Jessie Sun, Siyuan Hu, Wang Xiao-feng, Yong Chen, Ameya Nayate, Michael Wien, Douglas Martin, Lynn T. Singer, Jared Durieux, Chris A. Flask, Deanne Wilson CostelloarXiv 2023 · DOI · arXiv

This academic paper explores a significant hurdle in pediatric neurology: how to get high-quality brain images of irritable, non-sedated infants. By utilizing Magnetic Resonance Fingerprinting (MRF), the researchers demonstrate a path toward faster, more reliable, and quantitative neuroimaging for vulnerable populations.

TL;DR

Researchers successfully implemented a 5-minute "MR Fingerprinting" (MRF) scan that outperforms standard 28-minute clinical MRI protocols in imaging non-sedated neonates with prenatal opioid exposure. The technique provides motion-robust, high-resolution quantitative maps of brain tissue, offering a more efficient way to track infant brain development without the need for sedation.

What Problem Does This Paper Address?

Imaging the infant brain is a race against the "wiggles." Standard MRI requires subjects to remain perfectly still for long periods—a nearly impossible task for non-sedated babies, especially those with prenatal opioid exposure (POE) who are often more irritable and difficult to soothe.

Current clinical MRI has two major flaws in this context:

  1. Motion Sensitivity: Conventional "Cartesian" scanning is highly sensitive to movement, leading to high scan failure rates and the need for repeated sequences.
  2. Qualitative Data: Standard scans provide "weighted" images (T1-weighted or T2-weighted), which show relative contrast but lack absolute quantitative values. This makes it difficult to objectively compare brain maturation across different infants or longitudinal time points.

Methods

The study compared a traditional pediatric MRI protocol against a novel 3D MR Fingerprinting (MRF) sequence.

  • The MRF Approach: Unlike standard MRI, which acquires one type of image at a time, MRF uses a pseudorandom acquisition pattern (a "stack-of-spirals") to create unique signal "fingerprints" for different tissues. These are matched against a pre-calculated dictionary of signals based on the Bloch equations.
  • Synthetic Imaging: From a single 5-minute MRF scan, the team generated quantitative T1 and T2 maps, as well as "synthetic" clinical images (T1w, T2w, FLAIR, and DIR). Because these are derived from the same raw data, they are perfectly co-registered.
  • Evaluation: Four board-certified radiologists performed a "blinded" comparison of image quality between the standard MRI and the synthetic MRF images across 16 anatomical and myelination features.

Datasets

The study involved a cohort of 13 infants with prenatal opioid exposure, recruited between April 2021 and December 2022.

  • Demographics: 3 males, 10 females.
  • Age Range: 12 newborns (7–65 days old) and one 9-month-old infant for developmental comparison.
  • Context: This was a feasibility study conducted as part of the national ACT NOW: OBOE (Outcomes of Babies with Opioid Exposure) cohort.

Key Insights and Ideas

  • Inherent Motion Robustness: The "aha moment" here is that MRF’s non-Cartesian sampling and pattern-recognition-based reconstruction make it naturally resistant to motion artifacts. Even when babies moved, the MRF maps remained clear, while standard MRI scans often became unreadable.
  • Efficiency Overload: The MRF protocol took roughly 5 minutes, compared to the 28+ minutes required for the full clinical suite. Despite being much shorter, the MRF scan provided more data (quantitative maps + multiple synthetic contrasts).
  • Quantifying the "Un-quantifiable": The study captured the rapid myelination of the infant brain. They found that T1 and T2 values were significantly higher in babies under one month old compared to those aged 1–2 months, providing a concrete numerical metric for brain maturation that qualitative scans lack.

Key Results

  • Superior Image Quality: Radiologists rated MRF images as having significantly fewer artifacts and better visualization of anatomy. Specifically, 74.1% of MRF-T1w and 67.8% of MRF-T2w images received the highest quality grade, compared to 67.5% and 51.3% for standard MRI, respectively.
  • Scan Success: While 31% of standard MRI scans required repeats due to motion, no MRF scans needed to be repeated.
  • Developmental Sensitivity: The quantitative T1 and T2 values showed a statistically significant decrease as the infants aged from <1 month to 2 months ($p < 0.005$), matching expected patterns of brain maturation (increased myelin and decreased water content).

Limitations and Future Work

  • Sample Size: The study was limited to 13 subjects, partly due to recruitment challenges during the COVID-19 pandemic. Larger longitudinal studies are needed to establish growth curves for POE infants.
  • The "Artificial" Look: One radiologist noted an "artificial-ness" to the synthetic MRF images. This suggests a need for clinician "acclimatization" or further refinement of synthetic image rendering to match the "texture" clinicians are used to seeing.
  • Scan Order: In this study, MRF was always performed at the end of the session. Future work should investigate if performing MRF earlier (when the baby is potentially more settled) further improves results.
  • Motion Correction: While MRF is motion-robust, the authors suggest that integrating active motion correction techniques could provide even higher-quality maps in the future.

Summary updated

Abstract

Background: A noninvasive and sensitive imaging tool is needed to assess the fast-evolving baby brain. However, using MRI to study non-sedated babies faces roadblocks, including high scan failure rates due to subjects motion and the lack of quantitative measures for assessing potential developmental delays. This feasibility study explores whether MR Fingerprinting scans can provide motion-robust and quantitative brain tissue measurements for non-sedated infants with prenatal opioid exposure, presenting a viable alternative to clinical MR scans. Assessment: MRF image quality was compared to pediatric MRI scans using a fully crossed, multiple reader multiple case study. The quantitative T1 and T2 values were used to assess brain tissue changes between babies younger than one month and babies between one and two months. Statistical Tests: Generalized estimating equations (GEE) model was performed to test the significant difference of the T1 and T2 values from eight white matter regions of babies under one month and those are older. MRI and MRF image quality were assessed using Gwets second order auto-correlation coefficient (AC2) with its confidence levels. We used the Cochran-Mantel-Haenszel test to assess the difference in proportions between MRF and MRI for all features and stratified by the type of features. Results: In infants under one month of age, the T1 and T2 values are significantly higher (p&lt;0.005) compared to those between one and two months. A multiple-reader and multiple-case study showed superior image quality ratings in anatomical features from the MRF images than the MRI images. Conclusions: This study suggested that the MR Fingerprinting scans offer a motion-robust and efficient method for non-sedated infants, delivering superior image quality than clinical MRI scans and additionally providing quantitative measures to assess brain development.
BibTeX
@misc{ma2023motion,
  author = {Dan Ma and Chaitra Badve and Jessie Sun and Siyuan Hu and Wang Xiao-feng and Yong Chen and Ameya Nayate and Michael Wien and Douglas Martin and Lynn T. Singer and Jared Durieux and Chris A. Flask and Deanne Wilson Costello},
  title = {{Motion robust MR fingerprinting scan to image neonates with prenatal opioid exposure}},
  year = {2023},
  eprint = {2306.16656},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2306.16656},
}
Figure