Brain‐age estimation accuracy is significantly increased using multishell free‐water reconstruction

Although free‐water diffusion reconstruction for diffusion‐weighted imaging (DWI) data can be applied to both single‐shell and multishell data, recent finding in synthetic data suggests that the free‐water indices from single‐shell acquisition should be interpreted with care, as they are heavily influenced by initialization parameters and cannot discriminate between free‐water and mean diffusivity modifications. However, whether using a longer multishell acquisition protocol significantly improve reconstruction for real human MRI data is still an open question. In this study, we compare canonical diffusion tensor imaging (DTI), single‐shell and multishell free‐water imaging (FW) indices derived from a short, clinical compatible diffusion protocol (b = 500 s/mm², b = 1,000 s/mm², 32 directions each) on their power to predict brain age. Age was chosen as it is well‐known to be related to widespread modification of the white matter and because brain‐age estimation has recently been found to be relevant to several neurodegenerative diseases. We used a previously developed and validated data‐driven whole‐brain machine learning pipeline to directly compare the precision of brain‐age estimates in a sample of 89 healthy males between 20 and 85 years old. We found that multishell FW outperform DTI indices in estimating brain age and that multishell FW, even when using low (500 ms²) b‐values secondary shell, outperform single‐shell FW. Single‐shell FW led to lower brain‐age estimation accuracy even of canonical DTI indices, suggesting that single‐shell FW indices should be used with caution. For all considered reconstruction algorithms, the most discriminant indices were those measuring free diffusion of water in the white matter.     

Identifying a whole‐brain connectome‐based model in drug‐naïve Parkinson's disease for predicting motor impairment

Identifying a whole‐brain connectome‐based predictive model in drug‐naïve patients with Parkinson''s disease and verifying its predictions on drug‐managed patients would be useful in determining the intrinsic functional underpinnings of motor impairment and establishing general brain–behavior associations. In this study, we constructed a predictive model from the resting‐state functional data of 47 drug‐naïve patients by using a connectome‐based approach. This model was subsequently validated in 115 drug‐managed patients. The severity of motor impairment was assessed by calculating Unified Parkinson''s Disease Rating Scale Part III scores. The predictive performance of model was evaluated using the correlation coefficient (r _true) between predicted and observed scores. As a result, a connectome‐based model for predicting individual motor impairment in drug‐naïve patients was identified with significant performance (r _true = .845, p < .001, p _permu = .002). Two patterns of connection were identified according to correlations between connection strength and the severity of motor impairment. The negative motor‐impairment‐related network contained more within‐network connections in the motor, visual‐related, and default mode networks, whereas the positive motor‐impairment‐related network was constructed mostly with between‐network connections coupling the motor‐visual, motor‐limbic, and motor‐basal ganglia networks. Finally, this predictive model constructed around drug‐naïve patients was confirmed with significant predictive efficacy on drug‐managed patients (r = .209, p = .025), suggesting a generalizability in Parkinson''s disease patients under long‐term drug influence. In conclusion, this study identified a whole‐brain connectome‐based model that could predict the severity of motor impairment in Parkinson''s patients and furthers our understanding of the functional underpinnings of the disease.     

Intracranial and subcortical volumes in adolescents with early‐onset psychosis: A multisite mega‐analysis from the ENIGMA consortium

Early‐onset psychosis disorders are serious mental disorders arising before the age of 18 years. Here, we investigate the largest neuroimaging dataset, to date, of patients with early‐onset psychosis and healthy controls for differences in intracranial and subcortical brain volumes. The sample included 263 patients with early‐onset psychosis (mean age: 16.4 ± 1.4 years, mean illness duration: 1.5 ± 1.4 years, 39.2% female) and 359 healthy controls (mean age: 15.9 ± 1.7 years, 45.4% female) with magnetic resonance imaging data, pooled from 11 clinical cohorts. Patients were diagnosed with early‐onset schizophrenia (n = 183), affective psychosis (n = 39), or other psychotic disorders (n = 41). We used linear mixed‐effects models to investigate differences in intracranial and subcortical volumes across the patient sample, diagnostic subgroup and antipsychotic medication, relative to controls. We observed significantly lower intracranial (Cohen''s d = −0.39) and hippocampal (d = −0.25) volumes, and higher caudate (d = 0.25) and pallidum (d = 0.24) volumes in patients relative to controls. Intracranial volume was lower in both early‐onset schizophrenia (d = −0.34) and affective psychosis (d = −0.42), and early‐onset schizophrenia showed lower hippocampal (d = −0.24) and higher pallidum (d = 0.29) volumes. Patients who were currently treated with antipsychotic medication (n = 193) had significantly lower intracranial volume (d = −0.42). The findings demonstrate a similar pattern of brain alterations in early‐onset psychosis as previously reported in adult psychosis, but with notably low intracranial volume. The low intracranial volume suggests disrupted neurodevelopment in adolescent early‐onset psychosis.     

The additive impact of cardio‐metabolic disorders and psychiatric illnesses on accelerated brain aging

Severe mental illnesses (SMI) including major depressive disorder (MDD), bipolar disorder (BD), and schizophrenia spectrum disorder (SSD) elevate accelerated brain aging risks. Cardio‐metabolic disorders (CMD) are common comorbidities in SMI and negatively impact brain health. We validated a linear quantile regression index (QRI) approach against the machine learning “BrainAge” index in an independent SSD cohort (N = 206). We tested the direct and additive effects of SMI and CMD effects on accelerated brain aging in the N = 1,618 (604 M/1,014 F, average age = 63.53 ± 7.38) subjects with SMI and N = 11,849 (5,719 M/6,130 F; 64.42 ± 7.38) controls from the UK Biobank. Subjects were subdivided based on diagnostic status: SMI+/CMD+ (N = 665), SMI+/CMD− (N = 964), SMI−/CMD+ (N = 3,765), SMI−/CMD− (N = 8,083). SMI (F = 40.47, p = 2.06 × 10⁻¹⁰) and CMD (F = 24.69, p = 6.82 × 10⁻⁷) significantly, independently impacted whole‐brain QRI in SMI+. SSD had the largest effect (Cohen’s d = 1.42) then BD (d = 0.55), and MDD (d = 0.15). Hypertension had a significant effect on SMI+ (d = 0.19) and SMI− (d = 0.14). SMI effects were direct, independent of MD, and remained significant after correcting for effects of antipsychotic medications. Whole‐brain QRI was significantly (p < 10⁻¹⁶) associated with the volume of white matter hyperintensities (WMH). However, WMH did not show significant association with SMI and was driven by CMD, chiefly hypertension (p < 10⁻¹⁶). We used a simple and robust index, QRI, the demonstrate additive effect of SMI and CMD on accelerated brain aging. We showed a greater effect of psychiatric illnesses on QRI compared to cardio‐metabolic illness. Our findings suggest that subjects with SMI should be among the targets for interventions to protect against age‐related cognitive decline.     

Age‐specificity and generalization of behavior‐associated structural and functional networks and their relevance to behavioral domains

Behavior‐associated structural connectivity (SC) and resting‐state functional connectivity (rsFC) networks undergo various changes in aging. To study these changes, we proposed a continuous dimension where at one end networks generalize well across age groups in terms of behavioral predictions (age‐general) and at the other end, they predict behaviors well in a specific age group but fare poorly in another age group (age‐specific). We examined how age generalizability/specificity of multimodal behavioral associated brain networks varies across behavioral domains and imaging modalities. Prediction models consisting of SC and/or rsFC networks were trained to predict a diverse range of 75 behavioral outcomes in a young adult sample (N = 92). These models were then used to predict behavioral outcomes in unseen young (N = 60) and old (N = 60) subjects. As expected, behavioral prediction models derived from the young age group, produced more accurate predictions in the unseen young than old subjects. These behavioral predictions also differed significantly across behavioral domains, but not imaging modalities. Networks associated with cognitive functions, except for a few mostly relating to semantic knowledge, fell toward the age‐specific end of the spectrum (i.e., poor young‐to‐old generalizability). These findings suggest behavior‐associated brain networks are malleable to different degrees in aging; such malleability is partly determined by the nature of the behavior.     

Critical role of rhythms in prefrontal transcranial magnetic stimulation for depression: A randomized sham‐controlled study

Repetitive transcranial magnetic stimulation (rTMS) is an alternative treatment for depression, but the neural correlates of the treatment are currently inconclusive, which might be a limit of conventional analytical methods. The present study aimed to investigate the neurophysiological evidence and potential biomarkers for rTMS and intermittent theta burst stimulation (iTBS) treatment. A total of 61 treatment‐resistant depression patients were randomly assigned to receive prolonged iTBS (piTBS; N = 19), 10 Hz rTMS (N = 20), or sham stimulation (N = 22). Each participant went through a treatment phase with resting state electroencephalography (EEG) recordings before and after the treatment phase. The aftereffects of stimulation showed that theta‐alpha amplitude modulation frequency (f _am) was associated with piTBS_Responder, which involves repetitive bursts delivered in the theta frequency range, whereas alpha carrier frequency (f _c) was related to 10 Hz rTMS, which uses alpha rhythmic stimulation. In addition, theta‐alpha amplitude modulation frequency was positively correlated with piTBS antidepressant efficacy, whereas the alpha frequency was not associated with the 10 Hz rTMS clinical outcome. The present study showed that TMS stimulation effects might be lasting, with changes of brain oscillations associated with the delivered frequency. Additionally, theta‐alpha amplitude modulation frequency may be as a function of the degree of recovery in TRD with piTBS treatment and also a potential EEG‐based predictor of antidepressant efficacy of piTBS in the early treatment stage, that is, first 2 weeks.     

Automatic brain segmentation in preterm infants with post‐hemorrhagic hydrocephalus using 3D Bayesian U‐Net

Post‐hemorrhagic hydrocephalus (PHH) is a severe complication of intraventricular hemorrhage (IVH) in very preterm infants. PHH monitoring and treatment decisions rely heavily on manual and subjective two‐dimensional measurements of the ventricles. Automatic and reliable three‐dimensional (3D) measurements of the ventricles may provide a more accurate assessment of PHH, and lead to improved monitoring and treatment decisions. To accurately and efficiently obtain these 3D measurements, automatic segmentation of the ventricles can be explored. However, this segmentation is challenging due to the large ventricular anatomical shape variability in preterm infants diagnosed with PHH. This study aims to (a) propose a Bayesian U‐Net method using 3D spatial concrete dropout for automatic brain segmentation (with uncertainty assessment) of preterm infants with PHH; and (b) compare the Bayesian method to three reference methods: DenseNet, U‐Net, and ensemble learning using DenseNets and U‐Nets. A total of 41 T₂‐weighted MRIs from 27 preterm infants were manually segmented into lateral ventricles, external CSF, white and cortical gray matter, brainstem, and cerebellum. These segmentations were used as ground truth for model evaluation. All methods were trained and evaluated using 4‐fold cross‐validation and segmentation endpoints, with additional uncertainty endpoints for the Bayesian method. In the lateral ventricles, segmentation endpoint values for the DenseNet, U‐Net, ensemble learning, and Bayesian U‐Net methods were mean Dice score = 0.814 ± 0.213, 0.944 ± 0.041, 0.942 ± 0.042, and 0.948 ± 0.034 respectively. Uncertainty endpoint values for the Bayesian U‐Net were mean recall = 0.953 ± 0.037, mean  negative predictive value = 0.998 ± 0.005, mean accuracy = 0.906 ± 0.032, and mean AUC = 0.949 ± 0.031. To conclude, the Bayesian U‐Net showed the best segmentation results across all methods and provided accurate uncertainty maps. This method may be used in clinical practice for automatic brain segmentation of preterm infants with PHH, and lead to better PHH monitoring and more informed treatment decisions.     

Mutation‐related magnetization‐transfer,not axon density,drives white matter differences in premanifest Huntington disease: Evidence from in vivo ultra‐strong gradient MRI

White matter (WM) alterations have been observed in Huntington disease (HD) but their role in the disease‐pathophysiology remains unknown. We assessed WM changes in premanifest HD by exploiting ultra‐strong‐gradient magnetic resonance imaging (MRI). This allowed to separately quantify magnetization transfer ratio (MTR) and hindered and restricted diffusion‐weighted signal fractions, and assess how they drove WM microstructure differences between patients and controls. We used tractometry to investigate region‐specific alterations across callosal segments with well‐characterized early‐ and late‐myelinating axon populations, while brain‐wise differences were explored with tract‐based cluster analysis (TBCA). Behavioral measures were included to explore disease‐associated brain‐function relationships. We detected lower MTR in patients'' callosal rostrum (tractometry: p = .03; TBCA: p = .03), but higher MTR in their splenium (tractometry: p = .02). Importantly, patients'' mutation‐size and MTR were positively correlated (all p‐values < .01), indicating that MTR alterations may directly result from the mutation. Further, MTR was higher in younger, but lower in older patients relative to controls (p = .003), suggesting that MTR increases are detrimental later in the disease. Finally, patients showed higher restricted diffusion signal fraction (FR) from the composite hindered and restricted model of diffusion (CHARMED) in the cortico‐spinal tract (p = .03), which correlated positively with MTR in the posterior callosum (p = .033), potentially reflecting compensatory mechanisms. In summary, this first comprehensive, ultra‐strong gradient MRI study in HD provides novel evidence of mutation‐driven MTR alterations at the premanifest disease stage which may reflect neurodevelopmental changes in iron, myelin, or a combination of these.     

Sex differences in acute and long‐term brain recovery after concussion

Concussion is associated with acute disturbances in brain function and behavior, with potential long‐term effects on brain health. However, it is presently unclear whether there are sex differences in acute and long‐term brain recovery. In this study, magnetic resonance imaging (MRI) was used to scan 61 participants with sport‐related concussion (30 male, 31 female) longitudinally at acute injury, medical clearance to return to play (RTP), and 1‐year post‐RTP. A large cohort of 167 controls (80 male, 87 female) was also imaged. Each MRI session assessed cerebral blood flow (CBF), along with white matter fractional anisotropy (FA) and mean diffusivity (MD). For concussed athletes, the parameters were converted to difference scores relative to matched control subgroups, and partial least squares modeled the main and sex‐specific effects of concussion. Although male and female athletes did not differ in acute symptoms or time to RTP , all MRI measures showed significant sex differences during recovery. Males had greater reductions in occipital‐parietal CBF (mean difference and 95%CI: 9.97 ml/100 g/min, [4.84, 15.12] ml/100 g/min, z = 3.73) and increases in callosal MD (9.07 × 10⁻⁵, [−14.14, −3.60] × 10⁻⁵, z = −3.46), with greatest effects at 1‐year post‐RTP. In contrast, females had greater reductions in FA of the corona radiata (16.50 × 10⁻³, [−22.38, −11.08] × 10⁻³, z = −5.60), with greatest effects at RTP. These findings provide new insights into how the brain recovers after a concussion, showing sex differences in both the acute and chronic phases of injury.     

The interplay of resting and inhibitory control‐related theta‐band activity depends on age

Resting‐state neural activity plays an important role for cognitive control processes. Regarding response inhibition processes, an important facet of cognitive control, especially theta‐band activity has been the focus of research. Theoretical considerations suggest that the interrelation of resting and task‐related theta activity is subject to maturational effects. To investigate whether the relationship between resting theta activity and task‐related theta activity during a response inhibition task changes even in young age, we tested N = 166 healthy participants between 8 and 30 years of age. We found significant correlations between resting and inhibitory control‐related theta activity as well as behavioral inhibition performance. Importantly, these correlations were moderated by age. The moderation analysis revealed that higher resting theta activity was associated with stronger inhibition‐related theta activity in individuals above the age of ~10.7 years. The EEG beamforming analysis showed that this activity is associated with superior frontal region function (BA6). The correlation between resting and superior frontal response inhibition‐related theta activity became stronger with increasing age. A similar pattern was found for response inhibition performance, albeit only evident from the age of ~19.5 years. The results suggest that with increasing age, resting theta activity becomes increasingly important for processing the alarm/surprise signals in superior frontal brain regions during inhibitory control. Possible causes for these developmental changes are discussed.     

The sexual brain,genes, and cognition: A machine‐predicted brain sex score explains individual differences in cognitive intelligence and genetic influence in young children

Sex impacts the development of the brain and cognition differently across individuals. However, the literature on brain sex dimorphism in humans is mixed. We aim to investigate the biological underpinnings of the individual variability of sexual dimorphism in the brain and its impact on cognitive performance. To this end, we tested whether the individual difference in brain sex would be linked to that in cognitive performance that is influenced by genetic factors in prepubertal children (N = 9,658, ages 9–10 years old; the Adolescent Brain Cognitive Development study). To capture the interindividual variability of the brain, we estimated the probability of being male or female based on the brain morphometry and connectivity features using machine learning (herein called a brain sex score). The models accurately classified the biological sex with a test ROC–AUC of 93.32%. As a result, a greater brain sex score correlated significantly with greater intelligence (p _fdr < .001, ηp2 = .011–.034; adjusted for covariates) and higher cognitive genome‐wide polygenic scores (GPSs) (p _fdr < .001, ηp2 < .005). Structural equation models revealed that the GPS‐intelligence association was significantly modulated by the brain sex score, such that a brain with a higher maleness score (or a lower femaleness score) mediated a positive GPS effect on intelligence (indirect effects = .006–.009; p = .002–.022; sex‐stratified analysis). The finding of the sex modulatory effect on the gene–brain–cognition relationship presents a likely biological pathway to the individual and sex differences in the brain and cognitive performance in preadolescence.     

Loss of long‐term benefit from VIM‐DBS in essential tremor: A secondary analysis of repeated measurements

AimsDeep brain stimulation (DBS) in the ventral intermediate nucleus (Vim‐DBS) is the preferred surgical therapy for essential tremor (ET). Tolerance and disease progression are considered to be the two main reasons underlying the loss of long‐term efficacy of Vim‐DBS. This study aimed to explore whether Vim‐DBS shows long‐term loss of efficacy and to evaluate the reasons for this diminished efficacy from different aspects.MethodsIn a repeated‐measures meta‐analysis of 533 patients from 18 studies, Vim‐DBS efficacy was evaluated at ≤6 months, 7–12 months, 1–3 years, and ≥4 years. The primary outcomes were the score changes in different components of the Fahn‐Tolosa‐Marin Tremor Rating Scale (TRS; total score, motor score, hand‐function score, and activities of daily living [ADL] score). Secondary outcomes were the long‐term predictive factors.ResultsThe TRS total, motor, and ADL scores showed significant deterioration with disease progression (p = 0.002, p = 0.047, and p < 0.001, respectively), while the TRS total (p < 0.001), hand‐function (p = 0.036), and ADL (p = 0.004) scores indicated a significant long‐term reduction in DBS efficacy, although the motor subscore indicated no loss of efficacy. Hand‐function (p < 0.001) and ADL (p = 0.028) scores indicated DBS tolerance, while the TRS total and motor scores did not. Stimulation frequency and preoperative score were predictive factors for long‐term results.ConclusionThis study provides level 3a evidence that long‐term Vim‐DBS is effective in controlling motor symptoms without waning benefits. The efficacy reduction for hand function was caused by DBS tolerance, while that for ADL was caused by DBS tolerance and disease progression. More attention should be given to actual functional recovery rather than changes in motor scores in patients with ET.     

Stability test of canonical correlation analysis for studying brain‐behavior relationships: The effects of subject‐to‐variable ratios and correlation strengths

Canonical correlation analysis (CCA), a multivariate approach to identifying correlations between two sets of variables, is becoming increasingly popular in neuroimaging studies on brain‐behavior relationships. However, the CCA stability in neuroimaging applications has not been systematically investigated. Although it is known that the number of subjects should be greater than the number of variables due to the curse of dimensionality, it is unclear at what subject‐to‐variable ratios (SVR) and at what correlation strengths the CCA stability can be maintained. Here, we systematically assessed the CCA stability, in the context of investigating the relationship between the brain structural/functional imaging measures and the behavioral measures, by measuring the similarity of the first‐mode canonical variables across randomly sampled subgroups of subjects from a large set of 936 healthy subjects. Specifically, we tested how the CCA stability changes with SVR under two different brain‐behavior correlation strengths. The same tests were repeated using an independent data set (n = 700) for validation. The results confirmed that both SVR and correlation strength affect greatly the CCA stability—the CCA stability cannot be guaranteed if the SVR is not sufficiently high or the brain‐behavior relationship is not sufficiently strong. Based on our quantitative characterization of CCA stability, we provided a practical guideline to help correct interpretation of CCA results and proper applications of CCA in neuroimaging studies on brain‐behavior relationships.     

In vivo hippocampal subfield volumes in bipolar disorder—A mega‐analysis from The Enhancing Neuro Imaging Genetics through Meta‐Analysis Bipolar Disorder Working Group

The hippocampus consists of anatomically and functionally distinct subfields that may be differentially involved in the pathophysiology of bipolar disorder (BD). Here we, the Enhancing NeuroImaging Genetics through Meta‐Analysis Bipolar Disorder workinggroup, study hippocampal subfield volumetry in BD. T1‐weighted magnetic resonance imaging scans from 4,698 individuals (BD = 1,472, healthy controls [HC] = 3,226) from 23 sites worldwide were processed with FreeSurfer. We used linear mixed‐effects models and mega‐analysis to investigate differences in hippocampal subfield volumes between BD and HC, followed by analyses of clinical characteristics and medication use. BD showed significantly smaller volumes of the whole hippocampus (Cohen''s d = −0.20), cornu ammonis (CA)1 (d = −0.18), CA2/3 (d = −0.11), CA4 (d = −0.19), molecular layer (d = −0.21), granule cell layer of dentate gyrus (d = −0.21), hippocampal tail (d = −0.10), subiculum (d = −0.15), presubiculum (d = −0.18), and hippocampal amygdala transition area (d = −0.17) compared to HC. Lithium users did not show volume differences compared to HC, while non‐users did. Antipsychotics or antiepileptic use was associated with smaller volumes. In this largest study of hippocampal subfields in BD to date, we show widespread reductions in nine of 12 subfields studied. The associations were modulated by medication use and specifically the lack of differences between lithium users and HC supports a possible protective role of lithium in BD.     

Resting‐state magnetoencephalography source magnitude imaging with deep‐learning neural network for classification of symptomatic combat‐related mild traumatic brain injury

Combat‐related mild traumatic brain injury (cmTBI) is a leading cause of sustained physical, cognitive, emotional, and behavioral disabilities in Veterans and active‐duty military personnel. Accurate diagnosis of cmTBI is challenging since the symptom spectrum is broad and conventional neuroimaging techniques are insensitive to the underlying neuropathology. The present study developed a novel deep‐learning neural network method, 3D‐MEGNET, and applied it to resting‐state magnetoencephalography (rs‐MEG) source‐magnitude imaging data from 59 symptomatic cmTBI individuals and 42 combat‐deployed healthy controls (HCs). Analytic models of individual frequency bands and all bands together were tested. The All‐frequency model, which combined delta‐theta (1–7 Hz), alpha (8–12 Hz), beta (15–30 Hz), and gamma (30–80 Hz) frequency bands, outperformed models based on individual bands. The optimized 3D‐MEGNET method distinguished cmTBI individuals from HCs with excellent sensitivity (99.9 ± 0.38%) and specificity (98.9 ± 1.54%). Receiver‐operator‐characteristic curve analysis showed that diagnostic accuracy was 0.99. The gamma and delta‐theta band models outperformed alpha and beta band models. Among cmTBI individuals, but not controls, hyper delta‐theta and gamma‐band activity correlated with lower performance on neuropsychological tests, whereas hypo alpha and beta‐band activity also correlated with lower neuropsychological test performance. This study provides an integrated framework for condensing large source‐imaging variable sets into optimal combinations of regions and frequencies with high diagnostic accuracy and cognitive relevance in cmTBI. The all‐frequency model offered more discriminative power than each frequency‐band model alone. This approach offers an effective path for optimal characterization of behaviorally relevant neuroimaging features in neurological and psychiatric disorders.     

Distributed functional connectivity predicts neuropsychological test performance among older adults

Neuropsychological test is an essential tool in assessing cognitive and functional changes associated with late‐life neurocognitive disorders. Despite the utility of the neuropsychological test, the brain‐wide neural basis of the test performance remains unclear. Using the predictive modeling approach, we aimed to identify the optimal combination of functional connectivities that predicts neuropsychological test scores of novel individuals. Resting‐state functional connectivity and neuropsychological tests included in the OASIS‐3 dataset (n = 428) were used to train the predictive models, and the identified models were iteratively applied to the holdout internal test set (n = 216) and external test set (KSHAP, n = 151). We found that the connectivity‐based predicted score tracked the actual behavioral test scores (r = 0.08–0.44). The predictive models utilizing most of the connectivity features showed better accuracy than those composed of focal connectivity features, suggesting that its neural basis is largely distributed across multiple brain systems. The discriminant and clinical validity of the predictive models were further assessed. Our results suggest that late‐life neuropsychological test performance can be formally characterized with distributed connectome‐based predictive models, and further translational evidence is needed when developing theoretically valid and clinically incremental predictive models.     

Task‐related neural mechanisms of persecutory ideation in schizophrenia and community monozygotic twin‐pairs

Perceptions of spiteful behavior are common, distinct from rational fear, and may undergird persecutory ideation. To test this hypothesis and investigate neural mechanisms of persecutory ideation, we employed a novel economic social decision‐making task, the Minnesota Trust Game (MTG), during neuroimaging in patients with schizophrenia (n = 30) and community monozygotic (MZ) twins (n = 38; 19 pairs). We examined distinct forms of mistrust, task‐related brain activation and connectivity, and investigated relationships with persecutory ideation. We tested whether co‐twin discordance on these measurements was correlated to reflect a common source of underlying variance. Across samples persecutory ideation was associated with reduced trust only during the suspiciousness condition, which assessed spite sensitivity given partners had no monetary incentive to betray. Task‐based activation contrasts for specific forms of mistrust were limited and unrelated to persecutory ideation. However, task‐based connectivity contrasts revealed a dorsal cingulate anterior insula network sensitive to suspicious mistrust, a left frontal–parietal (lF‐P) network sensitive to rational mistrust, and a ventral medial/orbital prefrontal (vmPFC/OFC) network that was sensitive to the difference between these forms of mistrust (all p < .005). Higher persecutory ideation was predicted only by reduced connectivity between the vmPFC/OFC and lF‐P networks (p = .005), which was only observed when the intentions of the other player were relevant. Moreover, co‐twin differences in persecutory ideation predicted co‐twin differences in both spite sensitivity and in vmPFC/OFC–lF‐P connectivity. This work found that interconnectivity may be particularly important to the complex neurobiology underlying persecutory ideation, and that unique environmental variance causally linked persecutory ideation, decision‐making, and brain connectivity.     

Deep transfer learning of structural magnetic resonance imaging fused with blood parameters improves brain age prediction

Machine learning has been applied to neuroimaging data for estimating brain age and capturing early cognitive impairment in neurodegenerative diseases. Blood parameters like neurofilament light chain are associated with aging. In order to improve brain age predictive accuracy, we constructed a model based on both brain structural magnetic resonance imaging (sMRI) and blood parameters. Healthy subjects (n = 93; 37 males; aged 50–85 years) were recruited. A deep learning network was firstly pretrained on a large set of MRI scans (n = 1,481; 659 males; aged 50–85 years) downloaded from multiple open‐source datasets, to provide weights on our recruited dataset. Evaluating the network on the recruited dataset resulted in mean absolute error (MAE) of 4.91 years and a high correlation (r = .67, p <.001) against chronological age. The sMRI data were then combined with five blood biochemical indicators including GLU, TG, TC, ApoA1 and ApoB, and 9 dementia‐associated biomarkers including ApoE genotype, HCY, NFL, TREM2, Aβ40, Aβ42, T‐tau, TIMP1, and VLDLR to construct a bilinear fusion model, which achieved a more accurate prediction of brain age (MAE, 3.96 years; r = .76, p <.001). Notably, the fusion model achieved better improvement in the group of older subjects (70–85 years). Extracted attention maps of the network showed that amygdala, pallidum, and olfactory were effective for age estimation. Mediation analysis further showed that brain structural features and blood parameters provided independent and significant impact. The constructed age prediction model may have promising potential in evaluation of brain health based on MRI and blood parameters.     

Deep reasoning neural network analysis to predict language deficits from psychometry‐driven DWI connectome of young children with persistent language concerns

This study investigated whether current state‐of‐the‐art deep reasoning network analysis on psychometry‐driven diffusion tractography connectome can accurately predict expressive and receptive language scores in a cohort of young children with persistent language concerns (n = 31, age: 4.25 ± 2.38 years). A dilated convolutional neural network combined with a relational network (dilated CNN + RN) was trained to reason the nonlinear relationship between “dilated CNN features of language network” and “clinically acquired language score”. Three‐fold cross‐validation was then used to compare the Pearson correlation and mean absolute error (MAE) between dilated CNN + RN‐predicted and actual language scores. The dilated CNN + RN outperformed other methods providing the most significant correlation between predicted and actual scores (i.e., Pearson''s R/p‐value: 1.00/<.001 and .99/<.001 for expressive and receptive language scores, respectively) and yielding MAE: 0.28 and 0.28 for the same scores. The strength of the relationship suggests elevated probability in the prediction of both expressive and receptive language scores (i.e., 1.00 and 1.00, respectively). Specifically, sparse connectivity not only within the right precentral gyrus but also involving the right caudate had the strongest relationship between deficit in both the expressive and receptive language domains. Subsequent subgroup analyses inferred that the effectiveness of the dilated CNN + RN‐based prediction of language score(s) was independent of time interval (between MRI and language assessment) and age of MRI, suggesting that the dilated CNN + RN using psychometry‐driven diffusion tractography connectome may be useful for prediction of the presence of language disorder, and possibly provide a better understanding of the neurological mechanisms of language deficits in young children.     

White matter degeneration revealed by fiber‐specific analysis relates to recovery of hand function after stroke

Recent developments of higher‐order diffusion‐weighted imaging models have enabled the estimation of specific white matter fiber populations within a voxel, addressing limitations of traditional imaging markers of white matter integrity. We applied fixel based analysis (FBA) to investigate the evolution of fiber‐specific white matter changes in a prospective study of stroke patients and upper limb motor deficit over 1 year after stroke. We studied differences in fiber density and macrostructural changes in fiber cross‐section. Motor function was assessed by grip strength. We conducted a whole‐brain analysis of fixel metrics and predefined corticospinal tract (CST) region of interest in relation to changes in motor functions. In 30 stroke patients (mean age 62.3 years, SD ±16.9; median NIHSS 4, IQR 2–5), whole‐brain FBA revealed progressing loss of fiber density and cross‐section in the ipsilesional corticospinal tract and long‐range fiber tracts such as the superior longitudinal fascicle and trans‐callosal tracts extending towards contralesional white matter tracts. Lower FBA metrics measured at the brainstem section of the CST 1 month after stroke were significantly associated with lower grip strength 3 months (p = .009, adjusted R ² = 0.259) and 1 year (T4: p < .001, adj. R ² = 0.515) after stroke. Compared to FA, FBA metrics showed a comparably strong association with grip strength at later time points. Using FBA, we demonstrate progressive fiber‐specific white matter loss after stroke and association with functional motor outcome. Our results promote the application of fiber‐specific analysis to detect secondary neurodegeneration after stroke in relation to clinical recovery.