Differences in educational opportunity predict white matter development.

Coarse measures of socioeconomic status, such as parental income or parental education, have been linked to differences in white matter development. However, these measures do not provide insight into specific aspects of an individual's environment and how they relate to brain development. On the other hand, educational intervention studies have shown that changes in an individual's educational context can drive measurable changes in their white matter. These studies, however, rarely consider socioeconomic factors in their results. In the present study, we examined the unique relationship between educational opportunity and white matter development, when controlling other known socioeconomic factors. To explore this question, we leveraged the rich demographic and neuroimaging data available in the ABCD study, as well the unique data-crosswalk between ABCD and the Stanford Education Data Archive (SEDA). We find that educational opportunity is related to accelerated white matter development, even when accounting for other socioeconomic factors, and that this relationship is most pronounced in white matter tracts associated with academic skills. These results suggest that the school a child attends has a measurable relationship with brain development for years to come.

C9orf72 gene networks in the human brain correlate with cortical thickness in C9-FTD and implicate vulnerable cell types.

Introduction: A hexanucleotide repeat expansion (HRE) intronic to chromosome 9 open reading frame 72 (C9orf72) is recognized as the most common genetic cause of amyotrophic lateral sclerosis (ALS), frontotemporal dementia (FTD), and ALS-FTD. Identifying genes that show similar regional co-expression patterns to C9orf72 may help identify novel gene targets and biological mechanisms that mediate selective vulnerability to ALS and FTD pathogenesis. Methods: We leveraged mRNA expression data in healthy brain from the Allen Human Brain Atlas to evaluate C9orf72 co-expression patterns. To do this, we correlated average C9orf72 expression values in 51 regions across different anatomical divisions (cortex, subcortex, and cerebellum) with average gene expression values for 15,633 protein-coding genes, including 54 genes known to be associated with ALS, FTD, or ALS-FTD. We then performed imaging transcriptomic analyses to evaluate whether the identified C9orf72 co-expressed genes correlated with patterns of cortical thickness in symptomatic C9orf72 pathogenic HRE carriers (n = 19) compared to controls (n = 23). Lastly, we explored whether genes with significant C9orf72 imaging transcriptomic correlations (i.e., "C9orf72 imaging transcriptomic network") were enriched in specific cell populations in the brain and enriched for specific biological and molecular pathways. Results: A total of 2,120 genes showed an anatomical distribution of gene expression in the brain similar to C9orf72 and significantly correlated with patterns of cortical thickness in C9orf72 HRE carriers. This C9orf72 imaging transcriptomic network was differentially expressed in cell populations previously implicated in ALS and FTD, including layer 5b cells, cholinergic neurons in the spinal cord and brainstem and medium spiny neurons of the striatum, and was enriched for biological and molecular pathways associated with protein ubiquitination, autophagy, cellular response to DNA damage, endoplasmic reticulum to Golgi vesicle-mediated transport, among others. Conclusion: Considered together, we identified a network of C9orf72 associated genes that may influence selective regional and cell-type-specific vulnerabilities in ALS/FTD.

The University of California San Francisco Brain Metastases Stereotactic Radiosurgery (UCSF-BMSR) MRI Dataset.

Supplemental material is available for this article.

White matter and literacy: A dynamic system in flux.

Cross-sectional studies have linked differences in white matter tissue properties to reading skills. However, past studies have reported a range of, sometimes conflicting, results. Some studies suggest that white matter properties act as individual-level traits predictive of reading skill, whereas others suggest that reading skill and white matter develop as a function of an individual's educational experience. In the present study, we tested two hypotheses: a) that diffusion properties of the white matter reflect stable brain characteristics that relate to stable individual differences in reading ability or b) that white matter is a dynamic system, linked with learning over time. To answer these questions, we examined the relationship between white matter and reading in a five-year longitudinal dataset and a series of large-scale, single-observation, cross-sectional datasets (N = 14,249 total participants). We find that gains in reading skill correspond to longitudinal changes in the white matter. However, in the cross-sectional datasets, we find no evidence for the hypothesis that individual differences in white matter predict reading skill. These findings highlight the link between dynamic processes in the white matter and learning.

MR Thermometry during Transcranial MR Imaging-Guided Focused Ultrasound Procedures: A Review.

Interest in transcranial MR imaging-guided focused ultrasound procedures has recently grown. These incisionless procedures enable precise focal ablation of brain tissue using real-time monitoring by MR thermometry. This article will provide an updated review on clinically applicable technical underpinnings and considerations of proton resonance frequency MR thermometry, the most common clinically used MR thermometry sequence.

3D U-Net for automated detection of multiple sclerosis lesions: utility of transfer learning from other pathologies.

Background and purpose: Deep learning algorithms for segmentation of multiple sclerosis (MS) plaques generally require training on large datasets. This manuscript evaluates the effect of transfer learning from segmentation of another pathology to facilitate use of smaller MS-specific training datasets. That is, a model trained for detection of one type of pathology was re-trained to identify MS lesions and active demyelination. Materials and methods: In this retrospective study using MRI exams from 149 patients spanning 4/18/2014 to 7/8/2021, 3D convolutional neural networks were trained with a variable number of manually-segmented MS studies. Models were trained for FLAIR lesion segmentation at a single timepoint, new FLAIR lesion segmentation comparing two timepoints, and enhancing (actively demyelinating) lesion segmentation on T1 post-contrast imaging. Models were trained either de-novo or fine-tuned with transfer learning applied to a pre-existing model initially trained on non-MS data. Performance was evaluated with lesionwise sensitivity and positive predictive value (PPV). Results: For single timepoint FLAIR lesion segmentation with 10 training studies, a fine-tuned model demonstrated improved performance [lesionwise sensitivity 0.55 ± 0.02 (mean ± standard error), PPV 0.66 ± 0.02] compared to a de-novo model (sensitivity 0.49 ± 0.02, p = 0.001; PPV 0.32 ± 0.02, p < 0.001). For new lesion segmentation with 30 training studies and their prior comparisons, a fine-tuned model demonstrated similar sensitivity (0.49 ± 0.05) and significantly improved PPV (0.60 ± 0.05) compared to a de-novo model (sensitivity 0.51 ± 0.04, p = 0.437; PPV 0.43 ± 0.04, p = 0.002). For enhancement segmentation with 20 training studies, a fine-tuned model demonstrated significantly improved overall performance (sensitivity 0.74 ± 0.06, PPV 0.69 ± 0.05) compared to a de-novo model (sensitivity 0.44 ± 0.09, p = 0.001; PPV 0.37 ± 0.05, p = 0.001). Conclusion: By fine-tuning models trained for other disease pathologies with MS-specific data, competitive models identifying existing MS plaques, new MS plaques, and active demyelination can be built with substantially smaller datasets than would otherwise be required to train new models.

C9orf72 gene networks in the human brain correlate with cortical thickness in C9-FTD and implicate vulnerable cell types.

Introduction: A hexanucleotide repeat expansion (HRE) intronic to chromosome 9 open reading frame 72 (C9orf72) is recognized as the most common genetic cause of amyotrophic lateral sclerosis (ALS), frontotemporal dementia (FTD), and ALS-FTD. Identifying genes that show similar regional co-expression patterns to C9orf72 may help identify novel gene targets and biological mechanisms that mediate selective vulnerability to ALS and FTD pathogenesis. Methods: We leveraged mRNA expression data in healthy brain from the Allen Human Brain Atlas to evaluate C9orf72 co-expression patterns. To do this, we correlated average C9orf72 expression values in 51 regions across different anatomical divisions (cortex, subcortex, cerebellum) with average gene expression values for 15,633 protein-coding genes, including 50 genes known to be associated with ALS, FTD, or ALS-FTD. We then evaluated whether the identified C9orf72 co-expressed genes correlated with patterns of cortical thickness in symptomatic C9orf72 pathogenic HRE carriers (n=19). Lastly, we explored whether genes with significant C9orf72 radiogenomic correlations (i.e., 'C9orf72 gene network') were enriched in specific cell populations in the brain and enriched for specific biological and molecular pathways. Results: A total of 1,748 genes showed an anatomical distribution of gene expression in the brain similar to C9orf72 and significantly correlated with patterns of cortical thickness in C9orf72 HRE carriers. This C9orf72 gene network was differentially expressed in cell populations previously implicated in ALS and FTD, including layer 5b cells, cholinergic motor neurons in the spinal cord, and medium spiny neurons of the striatum, and was enriched for biological and molecular pathways associated with multiple neurotransmitter systems, protein ubiquitination, autophagy, and MAPK signaling, among others. Conclusions: Considered together, we identified a network of C9orf72-associated genes that may influence selective regional and cell-type-specific vulnerabilities in ALS/FTD.

Intracranial electrical stimulation of corticolimbic sites modulates arousal in humans.

BACKGROUND: Humans routinely shift their sleepiness and wakefulness levels in response to emotional factors. The diversity of emotional factors that modulates sleep-wake levels suggests that the ascending arousal network may be intimately linked with networks that mediate mood. Indeed, while animal studies have identified select limbic structures that play a role in sleep-wake regulation, the breadth of corticolimbic structures that directly modulates arousal in humans remains unknown. OBJECTIVE: We investigated whether select regional activation of the corticolimbic network through direct electrical stimulation can modulate sleep-wake levels in humans, as measured by subjective experience and behavior. METHODS: We performed intensive inpatient stimulation mapping in two human participants with treatment resistant depression, who underwent intracranial implantation with multi-site, bilateral depth electrodes. Stimulation responses of sleep-wake levels were measured by subjective surveys (i.e. Stanford Sleepiness Scale and visual-analog scale of energy) and a behavioral arousal score. Biomarker analyses of sleep-wake levels were performed by assessing spectral power features of resting-state electrophysiology. RESULTS: Our findings demonstrated three regions whereby direct stimulation modulated arousal, including the orbitofrontal cortex (OFC), subgenual cingulate (SGC), and, most robustly, ventral capsule (VC). Modulation of sleep-wake levels was frequency-specific: 100Hz OFC, SGC, and VC stimulation promoted wakefulness, whereas 1Hz OFC stimulation increased sleepiness. Sleep-wake levels were correlated with gamma activity across broad brain regions. CONCLUSIONS: Our findings provide evidence for the overlapping circuitry between arousal and mood regulation in humans. Furthermore, our findings open the door to new treatment targets and the consideration of therapeutic neurostimulation for sleep-wake disorders.

PRNP expression predicts imaging findings in sporadic Creutzfeldt-Jakob disease.

OBJECTIVE: We explored the relationship between regional PRNP expression from healthy brain tissue and patterns of increased and decreased diffusion and regional brain atrophy in patients with sporadic Creutzfeldt-Jakob disease (sCJD). METHODS: We used PRNP microarray data from 6 healthy adult brains from Allen Brain Institute and T1-weighted and diffusion-weighted MRIs from 34 patients diagnosed with sCJD and 30 age- and sex-matched healthy controls to construct partial correlation matrices across brain regions for specific measures of interest: PRNP expression, mean diffusivity, volume, cortical thickness, and local gyrification index, a measure of cortical folding. RESULTS: Regional patterns of PRNP expression in the healthy brain correlated with regional patterns of diffusion signal abnormalities and atrophy in sCJD. Among different measures of cortical morphology, regional patterns of local gyrification index in sCJD most strongly correlated with regional patterns of PRNP expression. At the vertex-wise level, different molecular subtypes of sCJD showed distinct regional correlations in local gyrification index across the cortex. Local gyrification index correlation patterns most closely matched patterns of PRNP expression in sCJD subtypes known to have greatest pathologic involvement of the cerebral cortex. INTERPRETATION: These results suggest that the specific genetic and molecular environment in which the prion protein is expressed confer variable vulnerability to misfolding across different brain regions that is reflected in patterns of imaging findings in sCJD. Further work in larger samples will be needed to determine whether these regional imaging patterns can serve as reliable markers of distinct disease subtypes to improve diagnosis and treatment targeting.

Radiogenomics of C9orf72 Expansion Carriers Reveals Global Transposable Element Derepression and Enables Prediction of Thalamic Atrophy and Clinical Impairment.

Hexanucleotide repeat expansion (HRE) within C9orf72 is the most common genetic cause of frontotemporal dementia (FTD). Thalamic atrophy occurs in both sporadic and familial FTD but is thought to distinctly affect HRE carriers. Separately, emerging evidence suggests widespread derepression of transposable elements (TEs) in the brain in several neurodegenerative diseases, including C9orf72 HRE-mediated FTD (C9-FTD). Whether TE activation can be measured in peripheral blood and how the reduction in peripheral C9orf72 expression observed in HRE carriers relates to atrophy and clinical impairment remain unknown. We used FreeSurfer software to assess the effects of C9orf72 HRE and clinical diagnosis (n = 78 individuals, male and female) on atrophy of thalamic nuclei. We also generated a novel, human, whole-blood RNA-sequencing dataset to determine the relationships among peripheral C9orf72 expression, TE activation, thalamic atrophy, and clinical severity (n = 114 individuals, male and female). We confirmed global thalamic atrophy and reduced C9orf72 expression in HRE carriers. Moreover, we identified disproportionate atrophy of the right mediodorsal lateral nucleus in HRE carriers and showed that C9orf72 expression associated with clinical severity, independent of thalamic atrophy. Strikingly, we found global peripheral activation of TEs, including the human endogenous LINE-1 element L1HS L1HS levels were associated with atrophy of multiple pulvinar nuclei, a thalamic region implicated in C9-FTD. Integration of peripheral transcriptomic and neuroimaging data from human HRE carriers revealed atrophy of specific thalamic nuclei, demonstrated that C9orf72 levels relate to clinical severity, and identified marked derepression of TEs, including L1HS, which predicted atrophy of FTD-relevant thalamic nuclei.SIGNIFICANCE STATEMENT Pathogenic repeat expansion in C9orf72 is the most frequent genetic cause of FTD and amyotrophic lateral sclerosis (ALS; C9-FTD/ALS). The clinical, neuroimaging, and pathologic features of C9-FTD/ALS are well characterized, whereas the intersections of transcriptomic dysregulation and brain structure remain largely unexplored. Herein, we used a novel radiogenomic approach to examine the relationship between peripheral blood transcriptomics and thalamic atrophy, a neuroimaging feature disproportionately impacted in C9-FTD/ALS. We confirmed reduction of C9orf72 in blood and found broad dysregulation of transposable elements-genetic elements typically repressed in the human genome-in symptomatic C9orf72 expansion carriers, which associated with atrophy of thalamic nuclei relevant to FTD. C9orf72 expression was also associated with clinical severity, suggesting that peripheral C9orf72 levels capture disease-relevant information.

Ordering Characteristics Predictive of Noncontrast CT Head Positivity in the Emergency Department.

RATIONALE AND OBJECTIVES: Recent decades have seen a steady increase in noncontrast head CT utilization in the emergency department with a concurrent rise in the practice of physician assistants (PAs) and nurse practitioners (NPs). The goal of this study was to identify ordering and patient characteristics predictive of positive noncontrast head CTs in the ED. We hypothesized NP/PAs would have lower positivity rates compared to physicians, suggestive of relative overutilization. MATERIALS AND METHODS: We retrospectively identified ED patients who underwent noncontrast head CTs at a single institution: a nonlevel 1 trauma center, during a 7-year period, recording examination positivity, ordering provider training/experience, and multiple additional ordering/patient attributes. Exam positivity was defined as any intracranial abnormality necessitating a change in acute management, such as acute hemorrhage, hydrocephalus, herniation, or worsening prior findings. RESULTS: 6624 patients met inclusion criteria. 4.6% (280/6107) of physician exams were positive while 3.7% (19/517) of NP/PA exams were positive; however, differences were not significant. Increasing provider experience was not associated with positivity. Attributes with increased positivity were patient age (p < 0.001), daytime exam (p < 0.05), and indications regarding malignancy (p < 0.001) or focal neurologic deficit (p = 0.001). Attributes with decreased positivity were indications of trauma (p < 0.001) or vertigo/dizziness (p < 0.05). CONCLUSION: We found no significant difference in rates of exam positivity between physicians and NP/PAs, even accounting for years of experience. This suggests increasing utilization of head CTs in the ED is not due to the increasing presence of NP/PAs, and may be reflective of general practice trends and clear diagnostic algorithms leading to head CT.

Tractography-based DBS lead repositioning improves outcome in refractory OCD and depression.

Deep brain stimulation (DBS) of the anterior limb of the internal capsule (ALIC) has been used to treat refractory obsessive-compulsive disorder (OCD) and depression, but outcomes are variable, with some patients not responding to this form of invasive neuromodulation. A lack of benefit in some patients may be due to suboptimal positioning of DBS leads. Recently, studies have suggested that specific white matter tracts within the ALIC are associated with improved outcomes. Here, we present the case of a patient who initially had a modest improvement in OCD and depressive symptoms after receiving DBS within the ALIC. Subsequently, he underwent unilateral DBS lead repositioning informed by tractography targeting the ventrolateral and medial prefrontal cortex's connection with the mediodorsal thalamus. In this patient, we also conducted post-implant and post-repositioning diffusion imaging and found that we could successfully perform tractography even with DBS leads in place. Following lead repositioning into tracts predictive of benefit, the patient reached responder criteria for his OCD, and his depression was remitted. This case illustrates that tractography can potentially be used in the evaluation and planning of lead repositioning to achieve therapeutic outcomes.

Longitudinal Assessment of Posttreatment Diffuse Glioma Tissue Volumes with Three-dimensional Convolutional Neural Networks.

Neural networks were trained for segmentation and longitudinal assessment of posttreatment diffuse glioma. A retrospective cohort (from January 2018 to December 2019) of 298 patients with diffuse glioma (mean age, 52 years ± 14 [SD]; 177 men; 152 patients with glioblastoma, 72 patients with astrocytoma, and 74 patients with oligodendroglioma) who underwent two consecutive multimodal MRI examinations were randomly selected into training (n = 198) and testing (n = 100) samples. A posttreatment tumor segmentation three-dimensional nnU-Net convolutional neural network with multichannel inputs (T1, T2, and T1 postcontrast and fluid-attenuated inversion recovery [FLAIR]) was trained to segment three multiclass tissue types (peritumoral edematous, infiltrated, or treatment-changed tissue [ED]; active tumor or enhancing tissue [AT]; and necrotic core). Separate longitudinal change nnU-Nets were trained on registered and subtracted FLAIR and T1 postlongitudinal images to localize and better quantify and classify changes in ED and AT. Segmentation Dice scores, volume similarities, and 95th percentile Hausdorff distances ranged from 0.72 to 0.89, 0.90 to 0.96, and 2.5 to 3.6 mm, respectively. Accuracy rates of the posttreatment tumor segmentation and longitudinal change networks being able to classify longitudinal changes in ED and AT as increased, decreased, or unchanged were 76%-79% and 90%-91%, respectively. The accuracy levels of the longitudinal change networks were not significantly different from those of three neuroradiologists (accuracy, 90%-92%; κ, 0.58-0.63; P > .05). The results of this study support the potential clinical value of artificial intelligence-based automated longitudinal assessment of posttreatment diffuse glioma. Keywords: MR Imaging, Neuro-Oncology, Neural Networks, CNS, Brain/Brain Stem, Segmentation, Quantification, Convolutional Neural Network (CNN) Supplemental material is available for this article. © RSNA, 2022.

Association of Genetic Variant Linked to Hemochromatosis With Brain Magnetic Resonance Imaging Measures of Iron and Movement Disorders.

Importance: Hereditary hemochromatosis (HH) is an autosomal recessive genetic disorder that leads to iron overload. Conflicting results from previous research has led some to believe the brain is spared the toxic effects of iron in HH. Objective: To test the association of the strongest genetic risk variant for HH on brainwide measures sensitive to iron deposition and the rates of movement disorders in a substantially larger sample than previous studies of its kind. Design, Setting, and Participants: This cross-sectional retrospective study included participants from the UK Biobank, a population-based sample. Genotype, health record, and neuroimaging data were collected from January 2006 to May 2021. Data analysis was conducted from January 2021 to April 2022. Disorders tested included movement disorders (International Statistical Classification of Diseases and Related Health Problems, Tenth Revision [ICD-10], codes G20-G26), abnormalities of gait and mobility (ICD-10 codes R26), and other disorders of the nervous system (ICD-10 codes G90-G99). Exposures: Homozygosity for p.C282Y, the largest known genetic risk factor for HH. Main Outcomes and Measures: T2-weighted and T2* signal intensity from brain magnetic resonance imaging scans, measures sensitive to iron deposition, and clinical diagnosis of neurological disorders. Results: The total cohort consisted of 488 288 individuals (264 719 female; ages 49-87 years, largely northern European ancestry), 2889 of whom were p.C282Y homozygotes. The neuroimaging analysis consisted of 836 individuals: 165 p.C282Y homozygotes (99 female) and 671 matched controls (399 female). A total of 206 individuals were excluded from analysis due to withdrawal of consent. Neuroimaging analysis showed that p.C282Y homozygosity was associated with decreased T2-weighted and T2* signal intensity in subcortical motor structures (basal ganglia, thalamus, red nucleus, and cerebellum; Cohen d >1) consistent with substantial iron deposition. Across the whole UK Biobank (2889 p.C282Y homozygotes, 485 399 controls), we found a significantly increased prevalence for movement disorders in male homozygotes (OR, 1.80; 95% CI, 1.28-2.55; P = .001) but not female individuals (OR, 1.09; 95% CI, 0.70-1.73; P = .69). Among the 31 p.C282Y male homozygotes with a movement disorder, only 10 had a concurrent HH diagnosis. Conclusions and Relevance: These findings indicate increased iron deposition in subcortical motor circuits in p.C282Y homozygotes and confirm an increased association with movement disorders in male homozygotes. Early treatment in HH effectively prevents the negative consequences of iron overload in the liver and heart. Our work suggests that screening for p.C282Y homozygosity in high-risk individuals also has the potential to reduce brain iron accumulation and to reduce the risk of movement disorders among male individuals who are homozygous for this mutation.

Deep Learning to Predict Neonatal and Infant Brain Age from Myelination on Brain MRI Scans.

Background Assessment of appropriate brain myelination on T1- and T2-weighted MRI scans is based on gestationally corrected age (GCA) and requires subjective visual inspection of the brain with knowledge of normal myelination milestones. Purpose To develop a convolutional neural network (CNN) capable of estimating neonatal and infant GCA based on brain myelination on MRI scans. Materials and methods In this retrospective study from one academic medical center, brain MRI scans of patients aged 0-25 months with reported normal myelination were consecutively collected between January 1995 and June 2019. The GCA at MRI was manually calculated. After exclusion criteria were applied, T1- and T2-weighted MRI scans were preprocessed with skull stripping, linear registration, z scoring for normalization, and downsampling. A three-dimensional regression CNN was trained to predict GCA using mean absolute error (MAE) as its loss function. Attention maps were calculated using layer-wise relevance propagation. Models were validated on an external test set from the National Institutes of Health (NIH). Model MAEs were compared using Kruskal-Wallis and Mann-Whitney tests. Results A total of 518 neonates and infants (mean GCA, 67 weeks ± 33 [SD], 56% male) was included, comprising 469 T1-, 438 T2-, and 389 T1- and T2-weighted studies. Across 10 runs, MAEs of T1-, T2-, and T1- and T2-weighted networks were 9.8 ± 2.3, 9.1 ± 1.9, and 7.7 ± 1.7 weeks, respectively. Attention map analysis demonstrated increased network attention to the cerebellum, posterior white matter, and basal ganglia signal in neonates with GCA of less than 40 weeks and the anterior white matter signal in infants with GCA of more than 120 weeks, corresponding to the known progression of myelination. The T1- and T2-weighted network tested on the external NIH test set had an MAE of 9.1 weeks, which was reduced to 5.9 weeks with further training using half the external test set (P < .001). Conclusion A three-dimensional convolutional neural network can predict the gestationally corrected age of neonates and infants aged 0-25 months based on brain myelination patterns on T1- and T2-weighted MRI scans. © RSNA, 2022 Online supplemental material is available for this article.

Interinstitutional Portability of a Deep Learning Brain MRI Lesion Segmentation Algorithm.

PURPOSE: To assess how well a brain MRI lesion segmentation algorithm trained at one institution performed at another institution, and to assess the effect of multi-institutional training datasets for mitigating performance loss. MATERIALS AND METHODS: In this retrospective study, a three-dimensional U-Net for brain MRI abnormality segmentation was trained on data from 293 patients from one institution (IN1) (median age, 54 years; 165 women; patients treated between 2008 and 2018) and tested on data from 51 patients from a second institution (IN2) (median age, 46 years; 27 women; patients treated between 2003 and 2019). The model was then trained on additional data from various sources: (a) 285 multi-institution brain tumor segmentations, (b) 198 IN2 brain tumor segmentations, and (c) 34 IN2 lesion segmentations from various brain pathologic conditions. All trained models were tested on IN1 and external IN2 test datasets, assessing segmentation performance using Dice coefficients. RESULTS: The U-Net accurately segmented brain MRI lesions across various pathologic conditions. Performance was lower when tested at an external institution (median Dice score, 0.70 [IN2] vs 0.76 [IN1]). Addition of 483 training cases of a single pathologic condition, including from IN2, did not raise performance (median Dice score, 0.72; P = .10). Addition of IN2 training data with heterogeneous pathologic features, representing only 10% (34 of 329) of total training data, increased performance to baseline (Dice score, 0.77; P < .001). This final model produced total lesion volumes with a high correlation to the reference standard (Spearman r = 0.98). CONCLUSION: For brain MRI lesion segmentation, adding a modest amount of relevant training data from an external institution to a previously trained model supported successful application of the model to this external institution.Keywords: Neural Networks, Brain/Brain Stem, Segmentation Supplemental material is available for this article. © RSNA, 2021.

Microstructural development from 9 to 14 years: Evidence from the ABCD Study.

During late childhood behavioral changes, such as increased risk-taking and emotional reactivity, have been associated with the maturation of cortico-cortico and cortico-subcortical circuits. Understanding microstructural changes in both white matter and subcortical regions may aid our understanding of how individual differences in these behaviors emerge. Restriction spectrum imaging (RSI) is a framework for modelling diffusion-weighted imaging that decomposes the diffusion signal from a voxel into hindered, restricted, and free compartments. This yields greater specificity than conventional methods of characterizing diffusion. Using RSI, we quantified voxelwise restricted diffusion across the brain and measured age associations in a large sample (n = 8086) from the Adolescent Brain and Cognitive Development (ABCD) study aged 9-14 years. Older participants showed a higher restricted signal fraction across the brain, with the largest associations in subcortical regions, particularly the basal ganglia and ventral diencephalon. Importantly, age associations varied with respect to the cytoarchitecture within white matter fiber tracts and subcortical structures, for example age associations differed across thalamic nuclei. This suggests that age-related changes may map onto specific cell populations or circuits and highlights the utility of voxelwise compared to ROI-wise analyses. Future analyses will aim to understand the relevance of this microstructural developmental for behavioral outcomes.