2368
Quantitative susceptibility mapping in the newborn brain with source separation: Comparison of preterm and full-term infants1Department of Radiology, Weill Cornell Medicine, New York, NY, United States, 2Department of Pediatrics, Weill Cornell Medicine, New York, NY, United States
Synopsis
Keywords: Neonatal, Quantitative Susceptibility mapping
Motivation: Estimating iron and myelin contents in the newborn brain can be used to access neurodevelopment, but is challenging with conventional quantitative susceptibility mapping.
Goal(s): To separate positive and negative susceptibilities in the newborn brain and investigate differences between preterm and full-term newborns.
Approach: A total of 22 full-term and 10 preterm newborns were studied using quantitative susceptibility mapping with R2*-based source separation. Mean susceptibilities within 10 regions-of-interests were compared between full-term and preterm newborns.
Results: Preterm brains showed less positive and negative susceptibilities, compared to the full-term brains.
Impact: This study suggests that positive and negative magnetic susceptibilities in the newborn brain may be estimated individually using quantitative susceptibility mapping with source separation and may be used to identify early deviations from normal neurodevelopment in preterm-born infants.
Introduction
Compared to full-term newborns, preterm newborns are at an increased risk for developmental disorders that may be characterized by iron deficiency or delayed myelination, potentially due to premature exposure to the extrauterine environment1. While conventional quantitative susceptibility mapping (QSM) can be used to assess iron and myelin contents in the newborn brain2-5, this technique can only present the average susceptibility within a voxel, and the opposing susceptibility sources of iron (positive) and myelin (negative) can cancel each other in the same voxel. This poses a challenge, especially in the newborn brain where one susceptibility source is not more dominant than the other. QSM with source separation has recently been demonstrated to examine the individual content of iron and myelin in the adult brain6-9. In this study, we applied QSM with source separation in the newborn brain to separate the contributions of iron and myelin and investigate differences between preterm and full-term newborns.Methods
This retrospective study was approved by our institutional review board. Fifty-six infants who underwent MRI within the first two months of life were initially included in this study. Infants were classified into full-term and preterm-born infants based on gestational age at birth (≥37 weeks vs <37 weeks). MRI scans were performed on 3 T GE (GE Healthcare, Waukesha, WI) or Siemens scanners (Siemens Healthcare, Erlangen, Germany). 3D multi-echo gradient echo (GRE) imaging data were acquired for QSM with the following scan parameters: TR = 49-81 ms, longest TE = 24-63 ms, matrix size = 224x192x32-416x320x56, and slice thickness = 2-3 mm. T2-weighted anatomical images were acquired using 3D fast spin echo imaging and were segmented using Draw-EM to generate the following 10 regions-of-interest (ROIs): cortical gray matter (CGM), white matter (WM), deep gray matter (DGM), caudate nucleus (CN), lentiform nucleus (LN), thalamus, and frontal, parietal, temporal, and occipital lobes. In QSM reconstruction, the variable-kernel sophisticated harmonic artifact reduction for phase data (V-SHARP) was used with a maximum sphere radius of 12 mm for background field removal10,11, and the morphological enabled dipole inversion (MEDI) algorithm was used for dipole inversion12,13. R2* was fitted using the fast mono-exponential fitting algorithm based on auto-regression on linear operations14. R2*-based source separation was performed with the relaxometric constant (Dr+,Dr-) of 137 Hz/ppm and the calibration parameter (α) of 1.919. Susceptibility maps were registered to anatomical images using ANTs (Fig. 1) and average values within each ROI were calculated. Differences in susceptibility and R2* between preterm and full-term newborns were evaluated using multiple linear regression, controlling for postmenstrual age (PMA) at MRI.Results
Of the 56 newborns, 19 were excluded due to insufficient image quality for analysis of either GRE or anatomical images (e.g., motion, blooming artifacts), and 5 were excluded due to severe brain structural abnormalities (e.g., hydrocephalus). As a result, a total of 22 full-term infants (mean PMA at MRI, 42.2 ± 2.4 weeks; 11 males) and 10 preterm (mean PMA at MRI, 37.5 ± 2.7 weeks; 6 males) newborns were studied. Compared to conventional QSM, QSM with source separation revealed somewhat increased regional contrasts, particularly in the basal ganglia on the positive susceptibility maps and in the posterior limb of the internal capsule on the negative susceptibility maps (Fig. 2). Figure 3 shows the mean susceptibilities and R2* of each ROI measured in all subjects. Compared to full-term newborns, preterm newborns exhibited significantly less positive susceptibilities in the CGM, LN, and frontal lobe, less negative susceptibilities in the CGM, frontal and parietal lobes, and lower R2* in the CGM, frontal and parietal lobes. In contrast, conventional QSM demonstrated substantially smaller magnitudes of susceptibility and detected a significant difference only in the LN between full-term and preterm newborns. Susceptibility and R2* maps of full-term and preterm infants with similar PMAs are shown in Fig. 4.Discussion & Conclusion
The separated positive and negative susceptibilities demonstrated greater differences between full-term and preterm newborns than the susceptibilities obtained from conventional QSM. Given the distinct contributions of iron and myelin contents to positive and negative susceptibilities, respectively, our findings suggest that preterm infants may experience deficiencies in both iron and myelin in the CGM and frontal lobe, iron deficiency in the LN, and delayed myelination in the parietal lobe. Further studies are warranted to investigate associations of the measured susceptibilities with neurodevelopmental outcomes. The newly developed source separation techniques in QSM reconstruction may be essential for a more accurate assessment of iron and myelin contents in the newborn brain and may help identify early biomarkers of neurodevelopmental disorders.Acknowledgements
R01HD100012, R01NS123576, R01NS095562, R01DK116126, R01AG080011, R01NS105144.References
1. Wang Y, Wu Y, Li T, Wang X, Zhu C. Iron metabolism and brain development in premature infants. Frontiers in physiology. 2019;10:463.
2. Zhang Y, Shi J, Wei H, Han V, Zhu WZ, Liu C. Neonate and infant brain development from birth to 2 years assessed using MRI-based quantitative susceptibility mapping. Neuroimage. 2019;185:349-360.
3. Ning N, Liu C, Wu P, et al. Spatiotemporal variations of magnetic susceptibility in the deep gray matter nuclei from 1 month to 6 years: A quantitative susceptibility mapping study. J Magn Reson Imaging. 2019;49:1600-1609.
4. Park SE, Mallak EJ, Wang Y, Nguyen TD, Zun Z. Comparison of brain quantitative susceptibility mapping between preterm and full-term newborns. ISMRM, 2023. p 5314.
5. Zun Z, Kapse K, Andersen N, Barnett S, Kapse A, Espinosa K, Quistorff J, Lopez C, Murnick J, Donofrio MT, Limperopoulos C. Quantitative susceptibility mapping in the infant brain diagnosed with congenital heart disease. ISMRM, 2021. p 2990.
6. Shin HG, Lee J, Yun YH, Yoo SH, Jang J, Oh SH, Nam Y, Jung S, Kim S, Fukunaga M, Kim W. χ-separation: Magnetic susceptibility source separation toward iron and myelin mapping in the brain. Neuroimage. 2021;240:118371.
7. Chen J, Gong NJ, Chaim KT, Otaduy MC, Liu C. Decompose quantitative susceptibility mapping (QSM) to sub-voxel diamagnetic and paramagnetic components based on gradient-echo MRI data. Neuroimage. 2021;242:118477.
8. Emmerich J, Bachert P, Ladd ME, Straub S. On the separation of susceptibility sources in quantitative susceptibility mapping: Theory and phantom validation with an in vivo application to multiple sclerosis lesions of different age. Journal of Magnetic Resonance. 2021;330:107033.
9. Dimov AV, Nguyen TD, Gillen KM, Marcille M, Spincemaille P, Pitt D, Gauthier SA, Wang Y. Susceptibility source separation from gradient echo data using magnitude decay modeling. Journal of Neuroimaging. 2022;32:852-859.
10. Schweser F, Deistung A, Lehr BW, Reichenbach JR. Quantitative imaging of intrinsic magnetic tissue properties using MRI signal phase: an approach to in vivo brain iron metabolism?. Neuroimage. 2011;54:2789-2807.
11. Özbay PS, Deistung A, Feng X, Nanz D, Reichenbach JR, Schweser F. A comprehensive numerical analysis of background phase correction with V‐SHARP. NMR in Biomedicine. 2017;30:e3550.
12. Liu T, Khalidov I, de Rochefort L, Spincemaille P, Liu J, Tsiouris AJ, Wang Y. A novel background field removal method for MRI using projection onto dipole fields. NMR in Biomedicine. 2011;24:1129-1136.
13. Liu J, Liu T, de Rochefort L, Ledoux J, Khalidov I, Chen W, Tsiouris AJ, Wisnieff C, Spincemaille P, Prince MR, Wang Y. Morphology enabled dipole inversion for quantitative susceptibility mapping using structural consistency between the magnitude image and the susceptibility map. Neuroimage. 2012;59:2560-2568.
14. Pei M, Nguyen TD, Thimmappa ND, Salustri C, Dong F, Cooper MA, Li J, Prince MR, Wang Y. Algorithm for fast monoexponential fitting based on auto‐regression on linear operations (ARLO) of data. Magnetic resonance in medicine. 2015;73:843-850.
Figures
