Characterisation of Children's Head Motion for Magnetic Resonance Imaging With and Without General Anaesthesia

Feasibility of single-shot multi-slice DENSE MRI at 7 T for strain tensor imaging in a paediatric population

Many children who have been treated for a posterior-fossa tumour experience neurocognitive problems after treatment with surgery, radiotherapy and/or chemotherapy, which significantly impact their quality of life. Knowledge about these underlying mechanisms is limited at this point. The displacement encoding with stimulated echoes (DENSE) sequence magnetic resonance imaging (MRI) at 7 T can be used to measure brain tissue pulsations, and provide information on both the blood vessels and microstructure simultaneously, which are potentially relevant parameters to assess these underlying mechanisms. A single-shot multi-slice DENSE sequence was used to obtain brain motion maps from which strain maps could be derived on a voxel-wise level. The robustness of this MRI sequence was studied using the root mean square displacement that was obtained during the registration in the analysis of the DENSE series. Although the paediatric participants exhibited noticeable head movement during the MR acquisition, good-quality strain maps were still obtained, displaying expected patterns similar to those seen in adults.

Whole-brain T2 mapping with radial sampling and retrospective motion correction at 3T

PURPOSE

Several barriers prevent the use of whole-brain T2 mapping in routine use despite increasing interest in this parameter. One of the main barriers is the long scan time resulting in patient discomfort and motion corrupted data. To address this challenge, a method for accurate whole-brain T2 mapping with a limited acquisition time and motion correction capabilities is investigated.

METHODS

A 3D radial multi-echo spin-echo sequence was implemented with optimized sampling trajectory enabling the estimation of intra-scan motion, subsequently used to correct the raw data. Motion corrected echo images are then reconstructed with linear subspace constrained reconstruction. Experiments were carried out on phantom and volunteers at 3T to evaluate the accuracy of the T2 estimation, the sensitivity to lesions and the efficiency of the correction on motion corrupted data.

RESULTS

Whole-brain T2 mapping acquired in less than 7 min enabled the depiction of lesions in the white matter with longer T2. Data retrospectively corrupted with typical motion traces of pediatric patients highly benefited from the motion correction by reducing the error in T2 estimates within the lesions. All datasets acquired on seven volunteers, with deliberate motion, also showed that motion corrupted T2 maps could be improved with the retrospective motion correction both at the voxel level and the structure level.

CONCLUSION

A whole-brain T2 mapping sequence with retrospective intra-scan motion correction and reasonable acquisition time is proposed. The method necessitates advanced iterative reconstruction strategies but no additional navigator, external device, or increased scan time is required.

Measuring and interpreting individual differences in fetal, infant, and toddler neurodevelopment

As scientists interested in fetal, infant, and toddler (FIT) neurodevelopment, our research questions often focus on how individual children differ in their neurodevelopment and the predictive value of those individual differences for long-term neural and behavioral outcomes. Measuring and interpreting individual differences in neurodevelopment can present challenges: Is there a "standard" way for the human brain to develop? How do the semantic, practical, or theoretical constraints that we place on studying "development" influence how we measure and interpret individual differences? While it is important to consider these questions across the lifespan, they are particularly relevant for conducting and interpreting research on individual differences in fetal, infant, and toddler neurodevelopment due to the rapid, profound, and heterogeneous changes happening during this period, which may be predictive of long-term outcomes. This article, therefore, has three goals: 1) to provide an overview about how individual differences in neurodevelopment are studied in the field of developmental cognitive neuroscience, 2) to identify challenges and considerations when studying individual differences in neurodevelopment, and 3) to discuss potential implications and solutions moving forward.

A neural network to create super-resolution MR from multiple 2D brain scans of pediatric patients

BACKGROUND

High-resolution (HR) 3D MR images provide detailed soft-tissue information that is useful in assessing long-term side-effects after treatment in childhood cancer survivors, such as morphological changes in brain structures. However, these images require long acquisition times, so routinely acquired follow-up images after treatment often consist of 2D low-resolution (LR) images (with thick slices in multiple planes).

PURPOSE

In this work, we present a super-resolution convolutional neural network, based on previous single-image MRI super-resolution work, that can reconstruct a HR image from 2D LR slices in multiple planes in order to facilitate the extraction of structural biomarkers from routine scans.

METHODS

A multilevel densely connected super-resolution convolutional neural network (mDCSRN) was adapted to take two perpendicular LR scans (e.g., coronal and axial) as tensors and reconstruct a 3D HR image. A training set of 90 HR T1 pediatric head scans from the Adolescent Brain Cognitive Development (ABCD) study was used, with 2D LR images simulated through a downsampling pipeline that introduces motion artifacts, blurring, and registration errors to make the LR scans more realistic to routinely acquired ones. The outputs of the model were compared against simple interpolation in two steps. First, the quality of the reconstructed HR images was assessed using the peak signal-to-noise ratio and structural similarity index compared to baseline. Second, the precision of structure segmentation (using the autocontouring software Limbus AI) in the reconstructed versus the baseline HR images was assessed using mean distance-to-agreement (mDTA) and 95% Hausdorff distance. Three datasets were used: 10 new ABCD images (dataset 1), 18 images from the Children's Brain Tumor Network (CBTN) study (dataset 2) and 6 "real-world" follow-up images of a pediatric head and neck cancer patient (dataset 3).

RESULTS

The proposed mDCSRN outperformed simple interpolation in terms of visual quality. Similarly, structure segmentations were closer to baseline images after 3D reconstruction. The mDTA improved to, on average (95% confidence interval), 0.7 (0.4-1.0) and 0.8 (0.7-0.9) mm for datasets 1 and 3 respectively, from the interpolation performance of 6.5 (3.6-9.5) and 1.2 (1.0-1.3) mm.

CONCLUSIONS

We demonstrate that deep learning methods can successfully reconstruct 3D HR images from 2D LR ones, potentially unlocking datasets for retrospective study and advancing research in the long-term effects of pediatric cancer. Our model outperforms standard interpolation, both in perceptual quality and for autocontouring. Further work is needed to validate it for additional structural analysis tasks.

Examining the reliability of brain age algorithms under varying degrees of participant motion

Brain age algorithms using data science and machine learning techniques show promise as biomarkers for neurodegenerative disorders and aging. However, head motion during MRI scanning may compromise image quality and influence brain age estimates. We examined the effects of motion on brain age predictions in adult participants with low, high, and no motion MRI scans (Original N = 148; Analytic N = 138). Five popular algorithms were tested: brainageR, DeepBrainNet, XGBoost, ENIGMA, and pyment. Evaluation metrics, intraclass correlations (ICCs), and Bland-Altman analyses assessed reliability across motion conditions. Linear mixed models quantified motion effects. Results demonstrated motion significantly impacted brain age estimates for some algorithms, with ICCs dropping as low as 0.609 and errors increasing up to 11.5 years for high motion scans. DeepBrainNet and pyment showed greatest robustness and reliability (ICCs = 0.956-0.965). XGBoost and brainageR had the largest errors (up to 13.5 RMSE) and bias with motion. Findings indicate motion artifacts influence brain age estimates in significant ways. Furthermore, our results suggest certain algorithms like DeepBrainNet and pyment may be preferable for deployment in populations where motion during MRI acquisition is likely. Further optimization and validation of brain age algorithms is critical to use brain age as a biomarker relevant for clinical outcomes.

A numerical human brain phantom for dynamic glucose-enhanced (DGE) MRI: On the influence of head motion at 3T

PURPOSE

Dynamic glucose-enhanced (DGE) MRI relates to a group of exchange-based MRI techniques where the uptake of glucose analogues is studied dynamically. However, motion artifacts can be mistaken for true DGE effects, while motion correction may alter true signal effects. The aim was to design a numerical human brain phantom to simulate a realistic DGE MRI protocol at 3T that can be used to assess the influence of head movement on the signal before and after retrospective motion correction.

METHODS

MPRAGE data from a tumor patient were used to simulate dynamic Z-spectra under the influence of motion. The DGE responses for different tissue types were simulated, creating a ground truth. Rigid head movement patterns were applied as well as physiological dilatation and pulsation of the lateral ventricles and head-motion-induced B0 -changes in presence of first-order shimming. The effect of retrospective motion correction was evaluated.

RESULTS

Motion artifacts similar to those previously reported for in vivo DGE data could be reproduced. Head movement of 1 mm translation and 1.5 degrees rotation led to a pseudo-DGE effect on the order of 1% signal change. B0 effects due to head motion altered DGE changes due to a shift in the water saturation spectrum. Pseudo DGE effects were partly reduced or enhanced by rigid motion correction depending on tissue location.

CONCLUSION

DGE MRI studies can be corrupted by motion artifacts. Designing post-processing methods using retrospective motion correction including B0 correction will be crucial for clinical implementation. The proposed phantom should be useful for evaluation and optimization of such techniques.