Field of View Normalization in Multi-Site Brain MRI

BOston Neonatal Brain Injury Data for Hypoxic Ischemic Encephalopathy (BONBID-HIE): I. MRI and Lesion Labeling

Hypoxic ischemic encephalopathy (HIE) is a brain injury that occurs in 1 ~ 5/1000 term neonates. Accurate identification and segmentation of HIE-related lesions in neonatal brain magnetic resonance images (MRIs) is the first step toward identifying high-risk patients, understanding neurological symptoms, evaluating treatment effects, and predicting outcomes. We release the first public dataset containing neonatal brain diffusion MRI and expert annotation of lesions from 133 patients diagnosed with HIE. HIE-related lesions in brain MRI are often diffuse (i.e., multi-focal), and small (over half the patients in our data having lesions occupying <1% of the brain volume (including ventricles)). Segmentation for HIE MRI data is remarkably different from, and arguably more challenging than, other segmentation tasks such as brain tumors with focal and relatively large lesions. We hope that this dataset can help fuel the development of MRI lesion segmentation methods for HIE and small diffuse lesions in general.

Deep learning of structural MRI predicts fluid, crystallized, and general intelligence

Can brain structure predict human intelligence? T1-weighted structural brain magnetic resonance images (sMRI) have been correlated with intelligence. However, the population-level association does not fully account for individual variability in intelligence. To address this, studies have emerged recently to predict individual subject's intelligence or neurocognitive scores. However, they are mostly on predicting fluid intelligence (the ability to solve new problems). Studies are lacking to predict crystallized intelligence (the ability to accumulate knowledge) or general intelligence (fluid and crystallized intelligence combined). This study tests whether deep learning of sMRI can predict an individual subject's verbal, comprehensive, and full-scale intelligence quotients (VIQ, PIQ, and FSIQ), which reflect fluid and crystallized intelligence. We performed a comprehensive set of 432 experiments, using different input image channels, six deep learning models, and two outcome settings, in 850 healthy and autistic subjects 6-64 years of age. Our findings indicate a statistically significant potential of T1-weighted sMRI in predicting intelligence, with a Pearson correlation exceeding 0.21 (p < 0.001). Interestingly, we observed that an increase in the complexity of deep learning models does not necessarily translate to higher accuracy in intelligence prediction. The interpretations of our 2D and 3D CNNs, based on GradCAM, align well with the Parieto-Frontal Integration Theory (P-FIT), reinforcing the theory's suggestion that human intelligence is a result of interactions among various brain regions, including the occipital, temporal, parietal, and frontal lobes. These promising results invite further studies and open new questions in the field.

aXonica: A support package for MRI based Neuroimaging

Magnetic Resonance Imaging (MRI) assists in studying the nervous system. MRI scans undergo significant processing before presenting the final images to medical practitioners. These processes are executed with ease due to excellent software pipelines. However, establishing software workstations is non-trivial and requires researchers in life sciences to be comfortable in downloading, installing, and scripting software that is non-user-friendly and may lack basic GUI. As researchers struggle with these skills, there is a dire need to develop software packages that can automatically install software pipelines speeding up building software workstations and laboratories. Previous solutions include NeuroDebian, BIDS Apps, Flywheel, QMENTA, Boutiques, Brainlife and Neurodesk. Overall, all these solutions complement each other. NeuroDebian covers neuroscience and has a wider scope, providing only 51 tools for MRI. Whereas, BIDS Apps is committed to the BIDS format, covering only 45 software related to MRI. Boutiques is more flexible, facilitating its pipelines to be easily installed as separate containers, validated, published, and executed. Whereas, both Flywheel and Qmenta are propriety, leaving four for users looking for 'free for use' tools, i.e., NeuroDebian, Brainlife, Neurodesk, and BIDS Apps. This paper presents an extensive survey of 317 tools published in MRI-based neuroimaging in the last ten years, along with 'aXonica,' an MRI-based neuroimaging support package that is unbiased towards any formatting standards and provides 130 applications, more than that of NeuroDebian (51), BIDS App (45), Flywheel (70), and Neurodesk (85). Using a technology stack that employs GUI as the front-end and shell scripted back-end, aXonica provides (i) 130 tools that span the entire MRI-based neuroimaging analysis, and allow the user to (ii) select the software of their choice, (iii) automatically resolve individual dependencies and (iv) installs them. Hence, aXonica can serve as an important resource for researchers and teachers working in the field of MRI-based Neuroimaging to (a) develop software workstations, and/or (b) install newer tools in their existing workstations.

A three-step, "brute-force" approach toward optimized affine spatial normalization

The first step in spatial normalization of magnetic resonance (MR) images commonly is an affine transformation, which may be vulnerable to image imperfections (such as inhomogeneities or "unusual" heads). Additionally, common software solutions use internal starting estimates to allow for a more efficient computation, which may pose a problem in datasets not conforming to these assumptions (such as those from children). In this technical note, three main questions were addressed: one, does the affine spatial normalization step implemented in SPM12 benefit from an initial inhomogeneity correction. Two, does using a complexity-reduced image version improve robustness when matching "unusual" images. And three, can a blind "brute-force" application of a wide range of parameter combinations improve the affine fit for unusual datasets in particular. A large database of 2081 image datasets was used, covering the full age range from birth to old age. All analyses were performed in Matlab. Results demonstrate that an initial removal of image inhomogeneities improved the affine fit particularly when more inhomogeneity was present. Further, using a complexity-reduced input image also improved the affine fit and was beneficial in younger children in particular. Finally, blindly exploring a very wide parameter space resulted in a better fit for the vast majority of subjects, but again particularly so in infants and young children. In summary, the suggested modifications were shown to improve the affine transformation in the large majority of datasets in general, and in children in particular. The changes can easily be implemented into SPM12.

Exploring approaches to tackle cross-domain challenges in brain medical image segmentation: a systematic review

Introduction

Brain medical image segmentation is a critical task in medical image processing, playing a significant role in the prediction and diagnosis of diseases such as stroke, Alzheimer's disease, and brain tumors. However, substantial distribution discrepancies among datasets from different sources arise due to the large inter-site discrepancy among different scanners, imaging protocols, and populations. This leads to cross-domain problems in practical applications. In recent years, numerous studies have been conducted to address the cross-domain problem in brain image segmentation.

Methods

This review adheres to the standards of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) for data processing and analysis. We retrieved relevant papers from PubMed, Web of Science, and IEEE databases from January 2018 to December 2023, extracting information about the medical domain, imaging modalities, methods for addressing cross-domain issues, experimental designs, and datasets from the selected papers. Moreover, we compared the performance of methods in stroke lesion segmentation, white matter segmentation and brain tumor segmentation.

Results

A total of 71 studies were included and analyzed in this review. The methods for tackling the cross-domain problem include Transfer Learning, Normalization, Unsupervised Learning, Transformer models, and Convolutional Neural Networks (CNNs). On the ATLAS dataset, domain-adaptive methods showed an overall improvement of ~3 percent in stroke lesion segmentation tasks compared to non-adaptive methods. However, given the diversity of datasets and experimental methodologies in current studies based on the methods for white matter segmentation tasks in MICCAI 2017 and those for brain tumor segmentation tasks in BraTS, it is challenging to intuitively compare the strengths and weaknesses of these methods.

Conclusion

Although various techniques have been applied to address the cross-domain problem in brain image segmentation, there is currently a lack of unified dataset collections and experimental standards. For instance, many studies are still based on n-fold cross-validation, while methods directly based on cross-validation across sites or datasets are relatively scarce. Furthermore, due to the diverse types of medical images in the field of brain segmentation, it is not straightforward to make simple and intuitive comparisons of performance. These challenges need to be addressed in future research.

All you need is data preparation: A systematic review of image harmonization techniques in Multi-center/device studies for medical support systems

BACKGROUND AND OBJECTIVES

Artificial intelligence (AI) models trained on multi-centric and multi-device studies can provide more robust insights and research findings compared to single-center studies. However, variability in acquisition protocols and equipment can introduce inconsistencies that hamper the effective pooling of multi-source datasets. This systematic review evaluates strategies for image harmonization, which standardizes appearances to enable reliable AI analysis of multi-source medical imaging.

METHODS

A literature search using PRISMA guidelines was conducted to identify relevant papers published between 2013 and 2023 analyzing multi-centric and multi-device medical imaging studies that utilized image harmonization approaches.

RESULTS

Common image harmonization techniques included grayscale normalization (improving classification accuracy by up to 24.42 %), resampling (increasing the percentage of robust radiomics features from 59.5 % to 89.25 %), and color normalization (enhancing AUC by up to 0.25 in external test sets). Initially, mathematical and statistical methods dominated, but machine and deep learning adoption has risen recently. Color imaging modalities like digital pathology and dermatology have remained prominent application areas, though harmonization efforts have expanded to diverse fields including radiology, nuclear medicine, and ultrasound imaging. In all the modalities covered by this review, image harmonization improved AI performance, with increasing of up to 24.42 % in classification accuracy and 47 % in segmentation Dice scores.

CONCLUSIONS

Continued progress in image harmonization represents a promising strategy for advancing healthcare by enabling large-scale, reliable analysis of integrated multi-source datasets using AI. Standardizing imaging data across clinical settings can help realize personalized, evidence-based care supported by data-driven technologies while mitigating biases associated with specific populations or acquisition protocols.

Maternal Childhood Abuse Versus Neglect Associated with Differential Patterns of Infant Brain Development

Severity of maternal childhood maltreatment has been associated with lower infant grey matter volume and amygdala volume during the first two years of life. A developing literature argues that effects of threat (abuse) and of deprivation (neglect) should be assessed separately because these distinct aspects of adversity may have different impacts on developmental outcomes. However, distinct effects of threat versus deprivation have not been assessed in relation to intergenerational effects of child maltreatment. The objective of this study was to separately assess the links of maternal childhood abuse and neglect with infant grey matter volume (GMV), white matter volume (WMV), amygdala and hippocampal volume. Participants included 57 mother-infant dyads. Mothers were assessed for childhood abuse and neglect using the Adverse Childhood Experiences (ACE) questionnaire in a sample enriched for childhood maltreatment. Between 4 and 24 months (M age = 12.28 months, SD = 5.99), under natural sleep, infants completed an MRI using a 3.0 T Siemens scanner. GMV, WMV, amygdala and hippocampal volumes were extracted via automated segmentation. Maternal history of neglect, but not abuse, was associated with lower infant GMV. Maternal history of abuse, but not neglect, interacted with age such that abuse was associated with smaller infant amygdala volume at older ages. Results are consistent with a threat versus deprivation framework, in which threat impacts limbic regions central to the stress response, whereas deprivation impacts areas more central to cognitive function. Further studies are needed to identify mechanisms contributing to these differential intergenerational associations of threat versus deprivation.

Linking maternal disrupted interaction and infant limbic volumes: The role of infant cortisol output

Despite a large animal literature documenting the role of low maternal nurturance and elevated glucocorticoid production on offspring limbic development, these pathways have not yet been assessed during human infancy. Informed by animal models, the present study examined whether 1) maternal disrupted interaction is related to infant cortisol levels, 2) infant cortisol levels are associated with infant limbic volumes, and 3) infant cortisol levels mediate associations between maternal disrupted interaction and infant limbic volumes. Participants included 57 mother-infant dyads. Infant saliva was measured at one time point before and two time points after the Still-Face Paradigm (SFP) at age 4 months. Five aspects of maternal disrupted interaction were coded during the SFP reunion episode. Between 4 and 25 months (M age = 11.74 months, SD = 6.12), under natural sleep, infants completed an MRI. Amygdala and hippocampal volumes were calculated via automated segmentation. Results indicated that 1) maternal disrupted interaction, and specifically disoriented interaction, with the infant was associated with higher infant salivary cortisol (AUCg) levels during the SFP, 2) higher infant AUCg was related to enlarged bilateral amygdala and hippocampal volumes, and 3) infant AUCg mediated the relation between maternal disrupted interaction and infant amygdala and hippocampal volumes. Findings are consistent with controlled animal studies and provide evidence of a link between increased cortisol levels and enlarged limbic volumes in human infants. Results further suggest that established interventions to decrease maternal disrupted interaction could impact both infant cortisol levels and infant limbic volumes.

Predictive value of diffusion MRI-based parametric response mapping for prognosis and treatment response in glioblastoma

BACKGROUND

Early detection of treatment response is important for the management of patients with malignant brain tumors such as glioblastoma to assure good quality of life in relation to therapeutic efficacy.

AIM

To investigate whether parametric response mapping (PRM) with diffusion MRI may provide prognostic information at an early stage of standard therapy for glioblastoma.

MATERIALS AND METHODS

This prospective study included 31 patients newly diagnosed with glioblastoma WHO grade IV, planned for primary standard postoperative treatment with radiotherapy 60Gy/30 fractions with concomitant and adjuvant Temozolomide. MRI follow-up including diffusion and perfusion weighting was performed at 3 T at start of postoperative chemoradiotherapy, three weeks into treatment, and then regularly until twelve months postoperatively. Regional mean diffusivity (MD) changes were analyzed voxel-wise using the PRM method (MD-PRM). At eight and twelve months postoperatively, after completion of standard treatment, patients were classified using conventional MRI and clinical evaluation as either having stable disease (SD, including partial response) or progressive disease (PD). It was assessed whether MD-PRM differed between patients having SD versus PD and whether it predicted the risk of disease progression (progression-free survival, PFS) or death (overall survival, OS). A subgroup analysis was performed that compared MD-PRM between SD and PD in patients only undergoing diagnostic biopsy. MGMT-promotor methylation status (O6-methylguanine-DNA methyltransferase) was registered and analyzed with respect to PFS, OS and MD-PRM.

RESULTS

Of the 31 patients analyzed: 21 were operated by resection and ten by diagnostic biopsy. At eight months, 19 patients had SD and twelve had PD. At twelve months, ten patients had SD and 20 had PD, out of which ten were deceased within twelve months and one was deceased without known tumor progression. Median PFS was nine months, and median OS was 17 months. Eleven patients had methylated MGMT-promotor, 16 were MGMT unmethylated, and four had unknown MGMT-status. MD-PRM did not significantly predict patients having SD versus PD neither at eight nor at twelve months. Patients with an above median MD-PRM reduction had a slightly longer PFS (P = 0.015) in Kaplan-Maier analysis, as well as a non-significantly longer OS (P = 0.099). In the subgroup of patients only undergoing biopsy, total MD-PRM change at three weeks was generally higher for patients with SD than for patients with PD at eight months, although no tests were performed. MGMT status strongly predicted both PFS and OS but not MD-PRM change.

CONCLUSION

MD-PRM at three weeks was not demonstrated to be predictive of treatment response, disease progression, or survival. Preliminary results suggested a higher predictive value in non-resected patients, although this needs to be evaluated in future studies.

Human-to-monkey transfer learning identifies the frontal white matter as a key determinant for predicting monkey brain age

The application of artificial intelligence (AI) to summarize a whole-brain magnetic resonance image (MRI) into an effective "brain age" metric can provide a holistic, individualized, and objective view of how the brain interacts with various factors (e.g., genetics and lifestyle) during aging. Brain age predictions using deep learning (DL) have been widely used to quantify the developmental status of human brains, but their wider application to serve biomedical purposes is under criticism for requiring large samples and complicated interpretability. Animal models, i.e., rhesus monkeys, have offered a unique lens to understand the human brain - being a species in which aging patterns are similar, for which environmental and lifestyle factors are more readily controlled. However, applying DL methods in animal models suffers from data insufficiency as the availability of animal brain MRIs is limited compared to many thousands of human MRIs. We showed that transfer learning can mitigate the sample size problem, where transferring the pre-trained AI models from 8,859 human brain MRIs improved monkey brain age estimation accuracy and stability. The highest accuracy and stability occurred when transferring the 3D ResNet [mean absolute error (MAE) = 1.83 years] and the 2D global-local transformer (MAE = 1.92 years) models. Our models identified the frontal white matter as the most important feature for monkey brain age predictions, which is consistent with previous histological findings. This first DL-based, anatomically interpretable, and adaptive brain age estimator could broaden the application of AI techniques to various animal or disease samples and widen opportunities for research in non-human primate brains across the lifespan.

Negative versus withdrawn maternal behavior: Differential associations with infant gray and white matter during the first 2 years of life

Distinct neural effects of threat versus deprivation emerge by childhood, but little data are available in infancy. Withdrawn versus negative parenting may represent dimensionalized indices of early deprivation versus early threat, but no studies have assessed neural correlates of withdrawn versus negative parenting in infancy. The objective of this study was to separately assess the links of maternal withdrawal and maternal negative/inappropriate interaction with infant gray matter volume (GMV), white matter volume (WMV), amygdala, and hippocampal volume. Participants included 57 mother-infant dyads. Withdrawn and negative/inappropriate aspects of maternal behavior were coded from the Still-Face Paradigm at four months infant age. Between 4 and 24 months (M age = 12.28 months, SD = 5.99), during natural sleep, infants completed an MRI using a 3.0 T Siemens scanner. GMV, WMV, amygdala, and hippocampal volumes were extracted via automated segmentation. Diffusion weighted imaging volumetric data were also generated for major white matter tracts. Maternal withdrawal was associated with lower infant GMV. Negative/inappropriate interaction was associated with lower overall WMV. Age did not moderate these effects. Maternal withdrawal was further associated with reduced right hippocampal volume at older ages. Exploratory analyses of white matter tracts found that negative/inappropriate maternal behavior was specifically associated with reduced volume in the ventral language network. Results suggest that quality of day-to-day parenting is related to infant brain volumes during the first two years of life, with distinct aspects of interaction associated with distinct neural effects.

BOston Neonatal Brain Injury Dataset for Hypoxic Ischemic Encephalopathy (BONBID-HIE): Part I. MRI and Manual Lesion Annotation

Hypoxic ischemic encephalopathy (HIE) is a brain injury that occurs in 1 ~ 5/1000 term neonates. Accurate identification and segmentation of HIE-related lesions in neonatal brain magnetic resonance images (MRIs) is the first step toward predicting prognosis, identifying high-risk patients, and evaluating treatment effects. It will lead to a more accurate estimation of prognosis, a better understanding of neurological symptoms, and a timely prediction of response to therapy. We release the first public dataset containing neonatal brain diffusion MRI and expert annotation of lesions from 133 patients diagnosed with HIE. HIE-related lesions in brain MRI are often diffuse (i.e., multi-focal), and small (over half the patients in our data having lesions occupying <1% of brain volume). Segmentation for HIE MRI data is remarkably different from, and arguably more challenging than, other segmentation tasks such as brain tumors with focal and relatively large lesions. We hope that this dataset can help fuel the development of MRI lesion segmentation methods for HIE and small diffuse lesions in general.

Maternal Childhood Maltreatment Is Associated With Lower Infant Gray Matter Volume and Amygdala Volume During the First Two Years of Life

Background

Childhood maltreatment affects approximately 25% of the world's population. Importantly, the children of mothers who have been maltreated are at increased risk of behavioral problems. Thus, one important priority is to identify child neurobiological processes associated with maternal childhood maltreatment (MCM) that might contribute to such intergenerational transmission. This study assessed the impact of MCM on infant gray and white matter volumes and infant amygdala and hippocampal volumes during the first 2 years of life.

Methods

Fifty-seven mothers with 4-month-old infants were assessed for MCM, using both the brief Adverse Childhood Experiences screening questionnaire and the more detailed Maltreatment and Abuse Chronology of Exposure scale. A total of 58% had experienced childhood maltreatment. Between 4 and 24 months (age in months: mean = 12.28, SD = 5.99), under natural sleep, infants completed a magnetic resonance imaging scan using a 3T Siemens scanner. Total brain volume, gray matter volume, white matter volume, and amygdala and hippocampal volumes were extracted via automated segmentation.

Results

MCM on the Adverse Childhood Experiences and Maltreatment and Abuse Chronology of Exposure scales were associated with lower infant total brain volume and gray matter volume, with no moderation by infant age. However, infant age moderated the association between MCM and right amygdala volume, such that MCM was associated with lower volume at older ages.

Conclusions

MCM is associated with alterations in infant brain volumes, calling for further identification of the prenatal and postnatal mechanisms contributing to such intergenerational transmission. Furthermore, the brief Adverse Childhood Experiences questionnaire predicted these alterations, suggesting the potential utility of early screening for infant risk.

Deep Relation Learning for Regression and Its Application to Brain Age Estimation

Most deep learning models for temporal regression directly output the estimation based on single input images, ignoring the relationships between different images. In this paper, we propose deep relation learning for regression, aiming to learn different relations between a pair of input images. Four non-linear relations are considered: "cumulative relation," "relative relation," "maximal relation" and "minimal relation." These four relations are learned simultaneously from one deep neural network which has two parts: feature extraction and relation regression. We use an efficient convolutional neural network to extract deep features from the pair of input images and apply a Transformer for relation learning. The proposed method is evaluated on a merged dataset with 6,049 subjects with ages of 0-97 years using 5-fold cross-validation for the task of brain age estimation. The experimental results have shown that the proposed method achieved a mean absolute error (MAE) of 2.38 years, which is lower than the MAEs of 8 other state-of-the-art algorithms with statistical significance (p<0.05) in paired T-test (two-side).

Tissue volume estimation and age prediction using rapid structural brain scans

The multicontrast EPImix sequence generates six contrasts, including a T1-weighted scan, in ~1 min. EPImix shows comparable diagnostic performance to conventional scans under qualitative clinical evaluation, and similarities in simple quantitative measures including contrast intensity. However, EPImix scans have not yet been compared to standard MRI scans using established quantitative measures. In this study, we compared conventional and EPImix-derived T1-weighted scans of 64 healthy participants using tissue volume estimates and predicted brain-age. All scans were pre-processed using the SPM12 DARTEL pipeline, generating measures of grey matter, white matter and cerebrospinal fluid volume. Brain-age was predicted using brainageR, a Gaussian Processes Regression model previously trained on a large sample of standard T1-weighted scans. Estimates of both global and voxel-wise tissue volume showed significantly similar results between standard and EPImix-derived T1-weighted scans. Brain-age estimates from both sequences were significantly correlated, although EPImix T1-weighted scans showed a systematic offset in predictions of chronological age. Supplementary analyses suggest that this is likely caused by the reduced field of view of EPImix scans, and the use of a brain-age model trained using conventional T1-weighted scans. However, this systematic error can be corrected using additional regression of T1-predicted brain-age onto EPImix-predicted brain-age. Finally, retest EPImix scans acquired for 10 participants demonstrated high test-retest reliability in all evaluated quantitative measurements. Quantitative analysis of EPImix scans has potential to reduce scanning time, increasing participant comfort and reducing cost, as well as to support automation of scanning, utilising active learning for faster and individually-tailored (neuro)imaging.

Comparison of cortical and subcortical structural segmentation methods in Alzheimer's disease: A statistical approach

BACKGROUND

Automated segmentation methods are developed to help with the segmentation of different brain areas. However, their reliability has yet to be fully investigated. To have a more comprehensive understanding of the distribution of changes in Alzheimer's disease (AD), as well as investigating the reliability of different segmentation methods, in this study we compared volumes of cortical and subcortical brain segments, using HIPS, volBrain, CAT and BrainSuite automated segmentation methods between AD, mild cognitive impairment (MCI) and healthy controls (HC).

METHODS

A total of 182 MRI images were taken from the minimal interval resonance imaging in Alzheimer's disease (MIRIAD; 22 AD and 22 HC) and the Alzheimer's disease neuroimaging initiative database (ADNI; 43 AD, 50 MCI and 45 HC) datasets. Statistical methods were used to compare different groups as well as the correlation between different methods.

RESULTS

The two methods of volBrain and CAT showed a strong correlation (p's < 0.035 Bonferroni corrected for multiple comparisons). The two methods, however, showed no significant correlation with BrainSuite (p's > 0.820 Bonferroni corrected). Furthermore, BrainSuite did not follow the same trend as the other three methods and only HIPS, volBrain and CAT showed strong conformity with the past literature with strong correlation with mini mental state examination (MMSE) scores.

CONCLUSION

Our results showed that automated segmentation methods HIPS, volBrain and CAT can be used in the classification of HC, AD and MCI. This is an indication that such methods can be used to inform researchers and clinicians of underlying mechanisms and progression of AD.

Study protocol: retrospectively mining multisite clinical data to presymptomatically predict seizure onset for individual patients with Sturge-Weber

INTRODUCTION

Secondary analysis of hospital-hosted clinical data can save time and cost compared with prospective clinical trials for neuroimaging biomarker development. We present such a study for Sturge-Weber syndrome (SWS), a rare neurovascular disorder that affects 1 in 20 000-50 000 newborns. Children with SWS are at risk for developing neurocognitive deficit by school age. A critical period for early intervention is before 2 years of age, but early diagnostic and prognostic biomarkers are lacking. We aim to retrospectively mine clinical data for SWS at two national centres to develop presymptomatic biomarkers.

METHODS AND ANALYSIS

We will retrospectively collect clinical, MRI and neurocognitive outcome data for patients with SWS who underwent brain MRI before 2 years of age at two national SWS care centres. Expert review of clinical records and MRI quality control will be used to refine the cohort. The merged multisite data will be used to develop algorithms for abnormality detection, lesion-symptom mapping to identify neural substrate and machine learning to predict individual outcomes (presence or absence of seizures) by 2 years of age. Presymptomatic treatment in 0-2 years and before seizure onset may delay or prevent the onset of seizures by 2 years of age, and thereby improve neurocognitive outcomes. The proposed work, if successful, will be one of the largest and most comprehensive multisite databases for the presymptomatic phase of this rare disease.

ETHICS AND DISSEMINATION

This study involves human participants and was approved by Boston Children's Hospital Institutional Review Board: IRB-P00014482 and IRB-P00025916 Johns Hopkins School of Medicine Institutional Review Board: NA_00043846. Participants gave informed consent to participate in the study before taking part. The Institutional Review Boards at Kennedy Krieger Institute and Boston Children's Hospital approval have been obtained at each site to retrospectively study this data. Results will be disseminated by presentations, publication and sharing of algorithms generated.

Diagnosis of early mild cognitive impairment using a multiobjective optimization algorithm based on T1-MRI data

Alzheimer's disease (AD) is the most prevalent form of dementia. The accurate diagnosis of AD, especially in the early phases is very important for timely intervention. It has been suggested that brain atrophy, as measured with structural magnetic resonance imaging (sMRI), can be an efficacy marker of neurodegeneration. While classification methods have been successful in diagnosis of AD, the performance of such methods have been very poor in diagnosis of those in early stages of mild cognitive impairment (EMCI). Therefore, in this study we investigated whether optimisation based on evolutionary algorithms (EA) can be an effective tool in diagnosis of EMCI as compared to cognitively normal participants (CNs). Structural MRI data for patients with EMCI (n = 54) and CN participants (n = 56) was extracted from Alzheimer's disease Neuroimaging Initiative (ADNI). Using three automatic brain segmentation methods, we extracted volumetric parameters as input to the optimisation algorithms. Our method achieved classification accuracy of greater than 93%. This accuracy level is higher than the previously suggested methods of classification of CN and EMCI using a single- or multiple modalities of imaging data. Our results show that with an effective optimisation method, a single modality of biomarkers can be enough to achieve a high classification accuracy.

Global-Local Transformer for Brain Age Estimation

Deep learning can provide rapid brain age estimation based on brain magnetic resonance imaging (MRI). However, most studies use one neural network to extract the global information from the whole input image, ignoring the local fine-grained details. In this paper, we propose a global-local transformer, which consists of a global-pathway to extract the global-context information from the whole input image and a local-pathway to extract the local fine-grained details from local patches. The fine-grained information from the local patches are fused with the global-context information by the attention mechanism, inspired by the transformer, to estimate the brain age. We evaluate the proposed method on 8 public datasets with 8,379 healthy brain MRIs with the age range of 0-97 years. 6 datasets are used for cross-validation and 2 datasets are used for evaluating the generality. Comparing with other state-of-the-art methods, the proposed global-local transformer reduces the mean absolute error of the estimated ages to 2.70 years and increases the correlation coefficient of the estimated age and the chronological age to 0.9853. In addition, our proposed method provides regional information of which local patches are most informative for brain age estimation. Our source code is available on: https://github.com/shengfly/global-local-transformer.

Multi-channel attention-fusion neural network for brain age estimation: Accuracy, generality, and interpretation with 16,705 healthy MRIs across lifespan

Brain age estimated by machine learning from T1-weighted magnetic resonance images (T1w MRIs) can reveal how brain disorders alter brain aging and can help in the early detection of such disorders. A fundamental step is to build an accurate age estimator from healthy brain MRIs. We focus on this step, and propose a framework to improve the accuracy, generality, and interpretation of age estimation in healthy brain MRIs. For accuracy, we used one of the largest sample sizes (N = 16,705). For each subject, our proposed algorithm first explicitly splits the T1w image, which has been commonly treated as a single-channel 3D image in other studies, into two 3D image channels representing contrast and morphometry information. We further proposed a "fusion-with-attention" deep learning convolutional neural network (FiA-Net) to learn how to best fuse the contrast and morphometry image channels. FiA-Net recognizes varying contributions across image channels at different brain anatomy and different feature layers. In contrast, multi-channel fusion does not exist for brain age estimation, and is mostly attention-free in other medical image analysis tasks (e.g., image synthesis, or segmentation), where treating channels equally may not be optimal. For generality, we used lifespan data 0-97 years of age for real-world utility; and we thoroughly tested FiA-Net for multi-site and multi-scanner generality by two phases of cross-validations in discovery and replication data, compared to most other studies with only one phase of cross-validation. For interpretation, we directly measured each artificial neuron's correlation with the chronological age, compared to other studies looking at the saliency of features where salient features may or may not predict age. Overall, FiA-Net achieved a mean absolute error (MAE) of 3.00 years and Pearson correlation r=0.9840 with known chronological ages in healthy brain MRIs 0-97 years of age, comparing favorably with state-of-the-art algorithms and studies for accuracy and generality across sites and datasets. We also provided interpretations on how different artificial neurons and real neuroanatomy contribute to the age estimation.

Voxelwise and Regional Brain Apparent Diffusion Coefficient Changes on MRI from Birth to 6 Years of Age

Background A framework for understanding rapid diffusion changes from 0 to 6 years of age is important in the detection of neurodevelopmental disorders. Purpose To quantify patterns of normal apparent diffusion coefficient (ADC) development from 0 to 6 years of age. Materials and Methods Previously constructed age-specific ADC atlases from 201 healthy full-term children (108 male; age range, 0-6 years) with MRI scans acquired from 2006 to 2013 at one large academic hospital were analyzed to quantify four patterns: ADC trajectory, rate of ADC change, age of ADC maturation, and hemispheric asymmetries of maturation ages. Patterns were quantified in whole-brain, segmented regional, and voxelwise levels by fitting a two-term exponential model. Hemispheric asymmetries in ADC maturation ages were assessed using t tests with Bonferroni correction. Results The posterior limb of the internal capsule (mean ADC: left hemisphere, 1.18 ×10³μm²/sec; right hemisphere, 1.17 ×10³μm²/sec), anterior limb of the internal capsule (left, 1.11 ×10³μm²/sec; right, 1.09 ×10³μm²/sec), vermis (1.26 ×10³μm²/sec), thalami (left, 1.17 ×10³μm²/sec; right, 1.15 ×10³μm²/sec), and basal ganglia (left, 1.26 ×10³μm²/sec; right, 1.23 ×10³μm²/sec) demonstrate low initial ADC values, indicating an earlier prenatal time course of development. ADC maturation was completed between 1.3 and 2.4 years of age, depending on the region. The vermis and left thalamus matured earliest (1.3 years). The frontolateral gray matter matured latest (right, 2.3 years; left, 2.4 years). ADC maturation occurred earlier in the left hemisphere (P < .001) in several regions, including the frontal (mean ± standard deviation) (left, 2.16 years ± 0.29; right, 2.19 years ± 0.31), temporal (left, 1.93 years ± 0.22; right, 1.99 years ± 0.22), and parietal (left, 1.92 years ± 0.30; right, 2.03 years ± 0.28) white matter. Maturation occurred earlier in the right hemisphere (P < .001) in several regions, including the thalami (left, 1.63 years ± 0.32; right, 1.45 years ± 0.33), basal ganglia (left, 1.79 years ± 0.31; right, 1.70 years ± 0.37), and hippocampi (left, 1.93 years ± 0.34; right, 1.78 years ± 0.33). Conclusion Normative apparent diffusion coefficient developmental patterns on diffusion-weighted MRI scans were quantified in children aged 0 to 6 years. This work provides knowledge about early brain development and may guide the detection of abnormal patterns of maturation. © RSNA, 2020 Online supplemental material is available for this article. See also the editorial by Rollins in this issue.

Maternal Dietary Intake of Omega-3 Fatty Acids Correlates Positively with Regional Brain Volumes in 1-Month-Old Term Infants

Maternal nutrition is an important factor for infant neurodevelopment. However, prior magnetic resonance imaging (MRI) studies on maternal nutrients and infant brain have focused mostly on preterm infants or on few specific nutrients and few specific brain regions. We present a first study in term-born infants, comprehensively correlating 73 maternal nutrients with infant brain morphometry at the regional (61 regions) and voxel (over 300 000 voxel) levels. Both maternal nutrition intake diaries and infant MRI were collected at 1 month of life (0.9 ± 0.5 months) for 92 term-born infants (among them, 54 infants were purely breastfed and 19 were breastfed most of the time). Intake of nutrients was assessed via standardized food frequency questionnaire. No nutrient was significantly correlated with any of the volumes of the 61 autosegmented brain regions. However, increased volumes within subregions of the frontal cortex and corpus callosum at the voxel level were positively correlated with maternal intake of omega-3 fatty acids, retinol (vitamin A) and vitamin B12, both with and without correction for postmenstrual age and sex (P < 0.05, q < 0.05 after false discovery rate correction). Omega-3 fatty acids remained significantly correlated with infant brain volumes after subsetting to the 54 infants who were exclusively breastfed, but retinol and vitamin B12 did not. This provides an impetus for future larger studies to better characterize the effect size of dietary variation and correlation with neurodevelopmental outcomes, which can lead to improved nutritional guidance during pregnancy and lactation.

Mining multi-site clinical data to develop machine learning MRI biomarkers: application to neonatal hypoxic ischemic encephalopathy

BACKGROUND

Secondary and retrospective use of hospital-hosted clinical data provides a time- and cost-efficient alternative to prospective clinical trials for biomarker development. This study aims to create a retrospective clinical dataset of Magnetic Resonance Images (MRI) and clinical records of neonatal hypoxic ischemic encephalopathy (HIE), from which clinically-relevant analytic algorithms can be developed for MRI-based HIE lesion detection and outcome prediction.

METHODS

This retrospective study will use clinical registries and big data informatics tools to build a multi-site dataset that contains structural and diffusion MRI, clinical information including hospital course, short-term outcomes (during infancy), and long-term outcomes (~ 2 years of age) for at least 300 patients from multiple hospitals.

DISCUSSION

Within machine learning frameworks, we will test whether the quantified deviation from our recently-developed normative brain atlases can detect abnormal regions and predict outcomes for individual patients as accurately as, or even more accurately, than human experts. Trial Registration Not applicable. This study protocol mines existing clinical data thus does not meet the ICMJE definition of a clinical trial that requires registration.

Deformable MRI-Ultrasound registration using correlation-based attribute matching for brain shift correction: Accuracy and generality in multi-site data

Intraoperative tissue deformation, known as brain shift, decreases the benefit of using preoperative images to guide neurosurgery. Non-rigid registration of preoperative magnetic resonance (MR) to intraoperative ultrasound (iUS) has been proposed as a means to compensate for brain shift. We focus on the initial registration from MR to predurotomy iUS. We present a method that builds on previous work to address the need for accuracy and generality of MR-iUS registration algorithms in multi-site clinical data. High-dimensional texture attributes were used instead of image intensities for image registration and the standard difference-based attribute matching was replaced with correlation-based attribute matching. A strategy that deals explicitly with the large field-of-view mismatch between MR and iUS images was proposed. Key parameters were optimized across independent MR-iUS brain tumor datasets acquired at 3 institutions, with a total of 43 tumor patients and 758 reference landmarks for evaluating the accuracy of the proposed algorithm. Despite differences in imaging protocols, patient demographics and landmark distributions, the algorithm is able to reduce landmark errors prior to registration in three data sets (5.37±4.27, 4.18±1.97 and 6.18±3.38 mm, respectively) to a consistently low level (2.28±0.71, 2.08±0.37 and 2.24±0.78 mm, respectively). This algorithm was tested against 15 other algorithms and it is competitive with the state-of-the-art on multiple datasets. We show that the algorithm has one of the lowest errors in all datasets (accuracy), and this is achieved while sticking to a fixed set of parameters for multi-site data (generality). In contrast, other algorithms/tools of similar performance need per-dataset parameter tuning (high accuracy but lower generality), and those that stick to fixed parameters have larger errors or inconsistent performance (generality but not the top accuracy). Landmark errors were further characterized according to brain regions and tumor types, a topic so far missing in the literature.