ABSTRACT
Individuals with congenital heart disease (CHD) have an increased risk of neurodevelopmental impairments. Given the hypothesized complexity linking genomics, atypical brain structure, cardiac diagnoses and their management, and neurodevelopmental outcomes, unsupervised methods may provide unique insight into neurodevelopmental variability in CHD. Using data from the Pediatric Cardiac Genomics Consortium Brain and Genes study, we identified data-driven subgroups of individuals with CHD from measures of brain structure. Using structural magnetic resonance imaging (MRI; N = 93; cortical thickness, cortical volume, and subcortical volume), we identified subgroups that differed primarily on cardiac anatomic lesion and language ability. In contrast, using diffusion MRI (N = 88; white matter connectivity strength), we identified subgroups that were characterized by differences in associations with rare genetic variants and visual-motor function. This work provides insight into the differential impacts of cardiac lesions and genomic variation on brain growth and architecture in patients with CHD, with potentially distinct effects on neurodevelopmental outcomes.
Subject(s)
Brain , Heart Defects, Congenital , Magnetic Resonance Imaging , Humans , Heart Defects, Congenital/pathology , Heart Defects, Congenital/diagnostic imaging , Heart Defects, Congenital/genetics , Female , Male , Child , Brain/diagnostic imaging , Brain/pathology , Adolescent , Young Adult , White Matter/diagnostic imaging , White Matter/pathology , Adult , Child, Preschool , Diffusion Magnetic Resonance Imaging , Neurodevelopmental Disorders/diagnostic imaging , Neurodevelopmental Disorders/pathology , Neurodevelopmental Disorders/geneticsABSTRACT
BACKGROUND: Limited serial neuroimaging studies use magnetic resonance imaging (MRI) to define the evolution of hypoxic-ischemic insults to the brain of term infants and encompass both the primary injury and its secondary impact on cerebral development. The optimal timing of MRI to fully evaluate the impact of hypoxic-ischemic encephalopathy on brain development and associated neurodevelopmental sequelae remains unknown. METHODS: Goals: (a) review literature related to serial neuroimaging in term infants with HIE; (b) describe pilot data in two infants with HIE treated with therapeutic hypothermia who had a brain injury at day 3-5 and underwent four additional MRIs over the next 12 weeks of life and developmental evaluation at 24 months of age. RESULTS: Early MRI defines primary injury on diffusion-weighted imaging, yet the full impact may not be fully apparent until after 1 month of life. CONCLUSION: The full impact of an ischemic injury on the neonatal brain may not be fully visible until several weeks after the initial insult. This suggests the benefit of obtaining later time points for MRI to fully define the extent of injury and its neurodevelopmental impact. IMPACT: Few studies inform the nature of the evolution of brain injury with hypothermia in HIE, limiting understanding of potential neuroprotection. MRI is the standard of care for prognosis in infants with HIE, however timing for optimal prognostic prediction remains unclear. Insights from MRI after the first week of life may assist in defining the full extent of brain injury and prognostic significance. A pilot study using five MRI timepoints up to 3 months of age, is presented. More data is required with a systematic evaluation of the impact of early brain injury on brain development in term infants with HIE following TH.
ABSTRACT
Rigid motion tracking is paramount in many medical imaging applications where movements need to be detected, corrected, or accounted for. Modern strategies rely on convolutional neural networks (CNN) and pose this problem as rigid registration. Yet, CNNs do not exploit natural symmetries in this task, as they are equivariant to translations (their outputs shift with their inputs) but not to rotations. Here we propose EquiTrack, the first method that uses recent steerable SE(3)-equivariant CNNs (E-CNN) for motion tracking. While steerable E-CNNs can extract corresponding features across different poses, testing them on noisy medical images reveals that they do not have enough learning capacity to learn noise invariance. Thus, we introduce a hybrid architecture that pairs a denoiser with an E-CNN to decouple the processing of anatomically irrelevant intensity features from the extraction of equivariant spatial features. Rigid transforms are then estimated in closed-form. EquiTrack outperforms state-of-the-art learning and optimisation methods for motion tracking in adult brain MRI and fetal MRI time series. Our code is available at https://github.com/BBillot/EquiTrack.