A machine learning investigation of volumetric and functional MRI abnormalities in adults born preterm

Relationship between abnormal intrinsic functional connectivity of subcortices and autism symptoms in high-functioning adults with autism spectrum disorder

PURPOSE

This study explores subcortices and their intrinsic functional connectivity (iFC) in autism spectrum disorder (ASD) adults and investigates their relationship with clinical severity.

METHODS

Resting-state functional magnetic resonance imaging (rs-fMRI) data were acquired from 74 ASD patients, and 63 gender and age-matched typically developing (TD) adults. Independent component analysis (ICA) was conducted to evaluate subcortical patterns of basal ganglia (BG) and thalamus. These two brain areas were treated as regions of interest to further calculate whole-brain FC. In addition, we employed multivariate machine learning to identify subcortices-based FC brain patterns and clinical scores to classify ASD adults from those TD subjects.

RESULTS

In ASD individuals, autism diagnostic observation schedule (ADOS) was negatively correlated with the BG network. Similarly, social responsiveness scale (SRS) was negatively correlated with the thalamus network. The BG-based iFC analysis revealed adults with ASD versus TD had lower FC, and its FC with the right medial temporal lobe (MTL), was positively correlated with SRS and ADOS separately. ASD could be predicted with a balanced accuracy of around 60.0 % using brain patterns and 84.7 % using clinical variables.

CONCLUSION

Our results revealed the abnormal subcortical iFC may be related to autism symptoms.

The past, current, and future of neonatal intensive care units with artificial intelligence: a systematic review

Machine learning and deep learning are two subsets of artificial intelligence that involve teaching computers to learn and make decisions from any sort of data. Most recent developments in artificial intelligence are coming from deep learning, which has proven revolutionary in almost all fields, from computer vision to health sciences. The effects of deep learning in medicine have changed the conventional ways of clinical application significantly. Although some sub-fields of medicine, such as pediatrics, have been relatively slow in receiving the critical benefits of deep learning, related research in pediatrics has started to accumulate to a significant level, too. Hence, in this paper, we review recently developed machine learning and deep learning-based solutions for neonatology applications. We systematically evaluate the roles of both classical machine learning and deep learning in neonatology applications, define the methodologies, including algorithmic developments, and describe the remaining challenges in the assessment of neonatal diseases by using PRISMA 2020 guidelines. To date, the primary areas of focus in neonatology regarding AI applications have included survival analysis, neuroimaging, analysis of vital parameters and biosignals, and retinopathy of prematurity diagnosis. We have categorically summarized 106 research articles from 1996 to 2022 and discussed their pros and cons, respectively. In this systematic review, we aimed to further enhance the comprehensiveness of the study. We also discuss possible directions for new AI models and the future of neonatology with the rising power of AI, suggesting roadmaps for the integration of AI into neonatal intensive care units.

Abnormal whole-brain voxelwise structure-function coupling and its association with cognitive dysfunction in patients with different cerebral small vessel disease burdens

Cerebral small vessel disease (CSVD) is a universal neurological disorder in older adults that occurs in connection with cognitive dysfunction and is a chief risk factor for dementia and stroke. While whole-brain voxelwise structural and functional abnormalities in CSVD have been heavily explored, the degree of structure-function coupling abnormality possible in patients with different CSVD burdens remains largely unknown. This study included 53 patients with severe CSVD burden (CSVD-s), 108 patients with mild CSVD burden (CSVD-m) and 76 healthy controls. A voxelwise coupling metric of low frequency fluctuations (ALFF) and voxel-based morphometry (VBM) was used to research the important differences in whole-brain structure-function coupling among groups. The correlations between ALFF/VBM decoupling and cognitive parameters in CSVD patients were then investigated. We found that compared with healthy controls, CSVD-s patients presented notably decreased ALFF/VBM coupling in the bilateral caudate nuclei and increased coupling in the right inferior temporal gyrus (ITG). In addition, compared with the CSVD-m group, the CSVD-s group demonstrated significantly decreased coupling in the bilateral caudate nuclei, right putamen and inferior frontal gyrus (IFG) and increased coupling in the left middle frontal gyrus and medial superior frontal gyrus. Notably, the ALFF/VBM decoupling values in the caudate, IFG and ITG not only showed significant correlations with attention and executive functions in CSVD patients but also prominently distinguished CSVD-s patients from CSVD-m patients and healthy controls in receiver operating characteristic curve research. Our discoveries demonstrated that decreased ALFF/VBM coupling in the basal ganglia and increased coupling in the frontotemporal lobes were connected with more severe burden and worse cognitive decline in CSVD patients. ALFF/VBM coupling might serve as a novel effective neuroimaging biomarker of CSVD burden and provide new insights into the pathophysiological mechanisms of the clinical development of CSVD.

Brain structural changes in blindness: a systematic review and an anatomical likelihood estimation (ALE) meta-analysis

In recent decades, numerous structural brain imaging studies investigated purported morphometric changes in early (EB) and late onset blindness (LB). The results of these studies have not yielded very consistent results, neither with respect to the type, nor to the anatomical locations of the brain morphometric alterations. To better characterize the effects of blindness on brain morphometry, we performed a systematic review and an Anatomical-Likelihood-Estimation (ALE) coordinate-based-meta-analysis of 65 eligible studies on brain structural changes in EB and LB, including 890 EB, 466 LB and 1257 sighted controls. Results revealed atrophic changes throughout the whole extent of the retino-geniculo-striate system in both EB and LB, whereas changes in areas beyond the occipital lobe occurred in EB only. We discuss the nature of some of the contradictory findings with respect to the used brain imaging methodologies and characteristics of the blind populations such as the onset, duration and cause of blindness. Future studies should aim for much larger sample sizes, eventually by merging data from different brain imaging centers using the same imaging sequences, opt for multimodal structural brain imaging, and go beyond a purely structural approach by combining functional with structural connectivity network analyses.

Long-lasting effects of very preterm birth on brain structure in adulthood: A systematic review and meta-analysis

Early life experiences, such as very preterm (VP) birth, can affect brain and cognitive development. Several prior studies investigated brain structure in adults born VP; synthesising these studies may help to provide a clearer understanding of long-term effects of VP birth on the brain. We systematically searched Medline and Embase for articles that investigated brain structure using MRI in adulthood in individuals born VP (<32 weeks' gestation) or with very low birth weight (VLBW; <1500 g), and controls born at term or with normal birth weight. In total, 77 studies met the review inclusion criteria, of which 28 studies were eligible for meta-analyses, including data from up to 797 VP/VLBW participants and 518 controls, aged 18-33 years. VP/VLBW adults exhibited volumetric, morphologic and microstructural alterations in subcortical and temporal cortical regions compared with controls, with pooled standardised mean differences up to - 1.0 (95% confidence interval: -1.2, -0.8). This study suggests there is a persisting neurological impact of VP birth, which may provide developmental neurobiological insights for adult cognition in high-risk populations.

Higher reliability and validity of Wavelet-ALFF of resting-state fMRI: From multicenter database and application to rTMS modulation

Amplitude of low-frequency fluctuation (ALFF) has been widely used for localization of abnormal activity at the single-voxel level in resting-state fMRI (RS-fMRI) studies. However, previous ALFF studies were based on fast Fourier transform (FFT-ALFF). Our recent study found that ALFF based on wavelet transform (Wavelet-ALFF) showed better sensitivity and reproducibility than FFT-ALFF. The current study aimed to test the reliability and validity of Wavelet-ALFF, and apply Wavelet-ALFF to investigate the modulation effect of repetitive transcranial magnetic stimulation (rTMS). The reliability and validity were assessed on multicenter RS-fMRI datasets under eyes closed (EC) and eyes open (EO) conditions (248 healthy participants in total). We then detected the sensitivity of Wavelet-ALFF using a rTMS modulation dataset (24 healthy participants). For each dataset, Wavelet-ALFF based on five mother wavelets (i.e., db2, bior4.4, morl, meyr and sym3) and FFT-ALFF were calculated in the conventional band and five frequency sub-bands. The results showed that the reliability of both inter-scanner and intra-scanner was higher with Wavelet-ALFF than with FFT-ALFF across multiple frequency bands, especially db2-ALFF in the higher frequency band slow-2 (0.1992-0.25 Hz). In terms of validity, the multicenter ECEO datasets showed that the effect sizes of Wavelet-ALFF with all mother wavelets (especially for db2-ALFF) were larger than those of FFT-ALFF across multiple frequency bands. Furthermore, Wavelet-ALFF detected a larger modulation effect than FFT-ALFF. Collectively, Wavelet db2-ALFF showed the best reliability and validity, suggesting that db2-ALFF may offer a powerful metric for inspecting regional spontaneous brain activities in future studies.

Adult outcome of preterm birth: Implications for neurodevelopmental theories of psychosis

Preterm birth is associated with an elevated risk of developmental and adult psychiatric disorders, including psychosis. In this review, we evaluate the implications of neurodevelopmental, cognitive, motor, and social sequelae of preterm birth for developing psychosis, with an emphasis on outcomes observed in adulthood. Abnormal brain development precipitated by early exposure to the extra-uterine environment, and exacerbated by neuroinflammation, neonatal brain injury, and genetic vulnerability, can result in alterations of brain structure and function persisting into adulthood. These alterations, including abnormal regional brain volumes and white matter macro- and micro-structure, can critically impair functional (e.g. frontoparietal and thalamocortical) network connectivity in a manner characteristic of psychotic illness. The resulting executive, social, and motor dysfunctions may constitute the basis for behavioural vulnerability ultimately giving rise to psychotic symptomatology. There are many pathways to psychosis, but elucidating more precisely the mechanisms whereby preterm birth increases risk may shed light on that route consequent upon early neurodevelopmental insult.

Multimodal classification of extremely preterm and term adolescents using the fusiform gyrus: A machine learning approach

OBJECTIVE

Extremely preterm birth has been associated with atypical visual and neural processing of faces, as well as differences in gray matter structure in visual processing areas relative to full-term peers. In particular, the right fusiform gyrus, a core visual area involved in face processing, has been shown to have structural and functional differences between preterm and full-term individuals from childhood through early adulthood. The current study used multiple neuroimaging modalities to build a machine learning model based on the right fusiform gyrus to classify extremely preterm birth status.

METHOD

Extremely preterm adolescents (n = 20) and full-term peers (n = 24) underwent structural and functional magnetic resonance imaging. Group differences in gray matter density, measured via voxel-based morphometry (VBM), and blood-oxygen level-dependent (BOLD) response to face stimuli were explored within the right fusiform. Using group difference clusters as seed regions, analyses investigating outgoing white matter streamlines, regional homogeneity, and functional connectivity during a face processing task and at rest were conducted. A data driven approach was utilized to determine the most discriminative combination of these features within a linear support vector machine classifier.

RESULTS

Group differences in two partially overlapping clusters emerged: one from the VBM analysis showing less density in the extremely preterm cohort and one from BOLD response to faces showing greater activation in the extremely preterm relative to full-term youth. A classifier fit to the data from the cluster identified in the BOLD analysis achieved an accuracy score of 88.64% when BOLD, gray matter density, regional homogeneity, and functional connectivity during the task and at rest were included. A classifier fit to the data from the cluster identified in the VBM analysis achieved an accuracy score of 95.45% when only BOLD, gray matter density, and regional homogeneity were included.

CONCLUSION

Consistent with previous findings, we observed neural differences in extremely preterm youth in an area that plays an important role in face processing. Multimodal analyses revealed differences in structure, function, and connectivity that, when taken together, accurately distinguish extremely preterm from full-term born youth. Our findings suggest a compensatory role of the fusiform where less dense gray matter is countered by increased local BOLD signal. Importantly, sub-threshold differences in many modalities within the same region were informative when distinguishing between extremely preterm and full-term youth.

Distinct effects of prematurity on MRI metrics of brain functional connectivity, activity, and structure: Univariate and multivariate analyses

Premature birth affects the developmental trajectory of the brain during a period of intense maturation with possible lifelong consequences. To better understand the effect of prematurity on brain structure and function, we performed blood‐oxygen‐level dependent (BOLD) and anatomical magnetic resonance imaging (MRI) at 40 weeks of postmenstrual age on 88 newborns with variable gestational age (GA) at birth and no evident radiological alterations. We extracted measures of resting‐state functional connectivity and activity in a set of 90 cortical and subcortical brain regions through the evaluation of BOLD correlations between regions and of fractional amplitude of low‐frequency fluctuation (fALFF) within regions, respectively. Anatomical information was acquired through the assessment of regional volumes. We performed univariate analyses on each metric to examine the association with GA at birth, the spatial distribution of the effects, and the consistency across metrics. Moreover, a data‐driven multivariate analysis (i.e., Machine Learning) framework exploited the high dimensionality of the data to assess the sensitivity of each metric to the effect of premature birth. Prematurity was associated with bidirectional alterations of functional connectivity and regional volume and, to a lesser extent, of fALFF. Notably, the effects of prematurity on functional connectivity were spatially diffuse, mainly within cortical regions, whereas effects on regional volume and fALFF were more focal, involving subcortical structures. While the two analytical approaches delivered consistent results, the multivariate analysis was more sensitive in capturing the complex pattern of prematurity effects. Future studies might apply multivariate frameworks to identify premature infants at risk of a negative neurodevelopmental outcome.

Advances in functional and diffusion neuroimaging research into the long-term consequences of very preterm birth

Very preterm birth (<32 weeks of gestation) has been associated with lifelong difficulties in a variety of neurocognitive functions. Magnetic resonance imaging (MRI) combined with advanced analytical approaches have been employed in order to increase our understanding of the neurodevelopmental problems that many very preterm born individuals face as they grow up. In this review, we will focus on two novel imaging techniques that have explored relationships between specific brain mechanisms and behavioural outcomes. These are functional MRI, which maps regional, time-varying changes in brain metabolism and diffusion-weighted MRI, which measures the displacement of water molecules in tissue and provides quantitative information about tissue microstructure. Identifying the neurobiological underpinning of the long-term sequelae associated with very preterm birth could inform the development and implementation of preventative interventions (before any cognitive problem emerges) and could facilitate the identification of behavioural targets for improving the life course outcomes of very preterm individuals.

Gray Matter Volume Differences Between More Versus Less Resilient Adults with Chronic Musculoskeletal Pain: A Voxel-based Morphology Study

Resilience, a personality construct that reflects capacities to persevere, maintain a positive outlook and/or thrive despite ongoing stressors, has emerged as an important focus of research on chronic pain (CP). Although behavior studies have found more resilient persons with CP experience less pain-related dysfunction than less resilient cohorts do, the presence and nature of associated brain structure differences has received scant attention. To address this gap, we examined gray matter volume (GMV) differences between more versus less resilient adults with chronic musculoskeletal pain. Participants (75 women, 43 men) were community-dwellers who reported ongoing musculoskeletal pain for at least three months. More (n = 57) and less (n = 61) resilient subgroups, respectively, were identified on the basis of scoring above and below median scores on two validated resilience questionnaires. Voxel-based morphology (VBM) undertaken to examine resilience subgroup differences in GMV indicated more resilient participants displayed significantly larger GMV in the (1) bilateral precuneus, (2) left superior and inferior parietal lobules, (3) orbital right middle frontal gyrus and medial right superior frontal gyrus, and (4) bilateral median cingulate and paracingulate gyri, even after controlling for subgroup differences on demographics and measures of pain-related distress. Together, results underscored the presence and nature of specific GMV differences underlying subjective reports of more versus less resilient responses to ongoing musculoskeletal pain.

Birth weight is associated with adolescent brain development: A multimodal imaging study in monozygotic twins

Previous research has shown that the prenatal environment, commonly indexed by birth weight (BW), is a predictor of morphological brain development. We previously showed in monozygotic (MZ) twins associations between BW and brain morphology that were independent of genetics. In the present study, we employed a longitudinal MZ twin design to investigate whether variations in prenatal environment (as indexed by discordance in BW) are associated with resting‐state functional connectivity (rs‐FC) and with structural connectivity. We focused on the limbic and default mode networks (DMNs), which are key regions for emotion regulation and internally generated thoughts, respectively. One hundred and six healthy adolescent MZ twins (53 pairs; 42% male pairs) followed longitudinally from birth underwent a magnetic resonance imaging session at age 15. Graph theoretical analysis was applied to rs‐FC measures. TrackVis was used to determine track count as an indicator of structural connectivity strength. Lower BW twins had less efficient limbic network connectivity as compared to their higher BW co‐twin, driven by differences in the efficiency of the right hippocampus and right amygdala. Lower BW male twins had fewer tracks connecting the right hippocampus and right amygdala as compared to their higher BW male co‐twin. There were no associations between BW and the DMN. These findings highlight the possible role of unique prenatal environmental influences in the later development of efficient spontaneous limbic network connections within healthy individuals, irrespective of DNA sequence or shared environment.

A machine learning investigation of volumetric and functional MRI abnormalities in adults born preterm

Imaging studies have characterized functional and structural brain abnormalities in adults after premature birth, but these investigations have mostly used univariate methods that do not account for hypothesized interdependencies between brain regions or quantify accuracy in identifying individuals. To overcome these limitations, we used multivariate machine learning to identify gray matter volume (GMV) and amplitude of low frequency fluctuations (ALFF) brain patterns that best classify young adults born very preterm/very low birth weight (VP/VLBW; n = 94) from those born full-term (FT; n = 92). We then compared the spatial maps of the structural and functional brain signatures and validated them by assessing associations with clinical birth history and basic cognitive variables. Premature birth could be predicted with a balanced accuracy of 80.7% using GMV and 77.4% using ALFF. GMV predictions were mediated by a pattern of subcortical and middle temporal reductions and volumetric increases of the lateral prefrontal, medial prefrontal, and superior temporal gyrus regions. ALFF predictions were characterized by a pattern including increases in the thalamus, pre- and post-central gyri, and parietal lobes, in addition to decreases in the superior temporal gyri bilaterally. Decision scores from each classification, assessing the degree to which an individual was classified as a VP/VLBW case, were predicted by the number of days in neonatal hospitalization and birth weight. ALFF decision scores also contributed to the prediction of general IQ, which highlighted their potential clinical significance. Combined, the results clarified previous research and suggested that primary subcortical and temporal damage may be accompanied by disrupted neurodevelopment of the cortex.