Parsing altered gray matter morphology of depression using a framework integrating the normative model and non-negative matrix factorization

Nature Communications, Journal Year: 2023, Volume and Issue: 14(1)

Published: July 8, 2023

Abstract The high inter-individual heterogeneity in individuals with depression limits neuroimaging studies case-control approaches to identify promising biomarkers for individualized clinical decision-making. We put forward a framework integrating the normative model and non-negative matrix factorization (NMF) quantitatively assess altered gray matter morphology from dimensional perspective. proposed parses into overlapping latent disease factors, assigns patients distinct factor compositions, thus preserving variability. identified four robust factors symptoms cognitive processes depression. In addition, we showed quantitative relationship between group-level morphological differences factors. Furthermore, this significantly predicted compositions of an independent dataset. provides approach resolve neuroanatomical

Language: Английский

---

Dimensional Neuroimaging Endophenotypes: Neurobiological Representations of Disease Heterogeneity Through Machine Learning

Biological Psychiatry, Journal Year: 2024, Volume and Issue: 96(7), P. 564 - 584

Published: May 6, 2024

Language: Английский

---

Stratifying ASD and characterizing the functional connectivity of subtypes in resting-state fMRI

Behavioural Brain Research, Journal Year: 2023, Volume and Issue: 449, P. 114458 - 114458

Published: April 29, 2023

Although stratifying autism spectrum disorder (ASD) into different subtypes is a common effort in the research field, few papers have characterized functional connectivity alterations of ASD subgroups classified by their clinical presentations.This case-control rs-fMRI study, based on large samples open database (Autism Brain Imaging Data Exchange, ABIDE). The rs-MRI data from n = 415 patients (males 357), and 574 typical development (TD) controls 410) were included. Clinical features extracted using each patient's Autism Diagnostic Interview-Revised (ADI-R) evaluation. Each subtype was local regional homogeneity (ReHo) for assessment, remote voxel-mirrored homotopic (VMHC) whole-brain connectivity, graph theoretical features. These identified imaging properties integrated to create machine learning model classifying data, an independent dataset used validate model.All participants Cluster-1 (patients with more severe impairment) Cluster-2 moderate according dimensional scores ADI-R. When compared TD group, demonstrated increased connection decreased widespread hyper- hypo-connectivity variations connectivity. quite similar group both But at level MCC-related connections specifically impaired Cluster-2. fused build model, which achieved ∼75% identifying (Cluster-1 accuracy 81.75%; 76.48%).The stratification presentations can help minimize disease heterogeneity highlight distinguished brain subtypes.

Language: Английский

---

Predicting the Symptom Severity in Autism Spectrum Disorder Based on EEG Metrics

IEEE Transactions on Neural Systems and Rehabilitation Engineering, Journal Year: 2022, Volume and Issue: 30, P. 1898 - 1907

Published: Jan. 1, 2022

Autism spectrum disorder (ASD) is associated with the impaired integrating and segregating of related information that expanded within large-scale brain network. The varying ASD symptom severities have been explored, relying on their behaviors activity, but how to effectively predict severity needs further exploration. In this study, we aim investigate whether could be predicted electroencephalography (EEG) metrics. Based a publicly available dataset, EEG networks were constructed, four types metrics calculated. Then, statistically compared network differences among children severities, i.e., high/low autism diagnostic observation schedule (ADOS) scores, as well typically developing (TD) children. Thereafter, utilized validate they facilitate prediction severity. results demonstrated both high-and low-scoring showed decreased long-range frontal-occipital connectivity, increased anterior frontal connectivity altered properties. Furthermore, found are significantly correlated ADOS combination can serve features current findings will expand our knowledge dysfunction in provide new for predicting

Language: Английский

---

Heterogeneous brain dynamic functional connectivity patterns in first‐episode drug‐naive patients with major depressive disorder

Human Brain Mapping, Journal Year: 2023, Volume and Issue: 44(8), P. 3112 - 3122

Published: March 15, 2023

Abstract It remains challenging to identify depression accurately due its biological heterogeneity. As people suffering from are associated with functional brain network alterations, we investigated subtypes of patients first‐episode drug‐naive (FEDN) based on characteristics. This study included data 91 FEDN and matched healthy individuals obtained the International Big‐Data Center for Depression Research. Twenty large‐scale connectivity networks were computed using group information guided independent component analysis. A multivariate unsupervised normative modeling method was used their networks, focusing individual‐level variability among quantifying deviations range. Two patient identified distinctive abnormal patterns, consisting 10 informative including default mode frontoparietal network. 16% belonged subtype I larger extreme normal range shorter illness duration, while 84% II weaker longer duration. Moreover, structural changes in more complex than patients. Compared controls, both increased decreased gray matter (GM) abnormalities widely distributed regions In contrast, most GM I. The patterns gleaned imaging can facilitate accurate identification FEDN‐MDD neurobiological

Language: Английский

---

Toward individual heterogeneity and neurobiological subtypes in attention-deficit/hyperactivity disorder

Journal of Pacific Rim Psychology, Journal Year: 2025, Volume and Issue: 19

Published: Jan. 1, 2025

Attention-deficit/hyperactivity disorder (ADHD) is a biologically and clinically heterogeneous neurodevelopmental condition, which hinders the identification of rooted evidence for treatment choices clinical predictions. Identifying brain-based homogenous ADHD subtypes with neuroimaging data to reduce this heterogeneity promising elucidating specific neural mechanisms underlying complex presentations, may enable development personalized treatments precise therapeutic targets. In review, we first discuss large individual differences among patients indicated by findings from both large-scale group-level studies individual-level studies, motivated new efforts discover neurobiological subtypes. Next, review recent research on neuroimaging-based in terms three aspects: sample selection, subtyping methodology (i.e., features, algorithms, validation strategies), subtype findings. Eleven utilizing multiple single modalities or multimodal were identified. Through diverse features approaches, current have revealed range different characterized distinct profiles, providing important insight into nature ADHD. Despite progress, most still little biological relevance, limited utility, generalizability, slowing down pace their translation. We highlight several crucial considerations overcome these challenges contribute more useful reproducible identification. With increasing access datasets, deliberate features/methods adequate strategies, believe that could be used inform treatments, thereby advancing practice towards precision psychiatry.

Language: Английский

---

Parsing the heterogeneity of brain structure and function in male children with autism spectrum disorder: a multimodal MRI study

Brain Imaging and Behavior, Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 18, 2025

Language: Английский

---

Neurofind: using deep learning to make individualised inferences in brain-based disorders

Translational Psychiatry, Journal Year: 2025, Volume and Issue: 15(1)

Published: Feb. 27, 2025

Abstract Within precision psychiatry, there is a growing interest in normative models given their ability to parse heterogeneity. While they are intuitive and informative, the technical expertise resources required develop may not be accessible most researchers. Here we present Neurofind, new freely available tool that bridges this gap by wrapping sound previously tested methods on data harmonisation advanced into web-based platform requires minimal input from user. We explain how Neurofind was developed, use website four simple steps ( www.neurofind.ai ), provide exemplar applications. takes as structural MRI images outputs two main metrics derived independent models: (1) Outlier Index Score, deviation score brain morphology, (2) Brain Age, predicted age based an individual’s morphometry. The trained 3362 of healthy controls aged 20–80 publicly datasets. volume 101 cortical subcortical regions extracted modelled with adversarial autoencoder for index model support vector regression model. To illustrate potential applications, applied 364 three datasets patients diagnosed Alzheimer’s disease schizophrenia. In disease, 55.2% had very extreme Scores, mostly driven larger deviations temporal-limbic structures ventricles. Patients were also homogeneous deviated norm. Conversely, only 30.1% schizophrenia outliers, due hippocampus pallidum, tended more heterogeneous than controls. Both groups showed signs accelerated ageing.

Language: Английский

---

Normative modeling of brain morphometry in self-limited epilepsy with centrotemporal spikes

Cerebral Cortex, Journal Year: 2025, Volume and Issue: 35(3)

Published: March 1, 2025

Self-limited epilepsy with centrotemporal spikes is the most common pediatric epilepsy, characterized by an age-dependent onset that typically arises during childhood brain development and followed remission at puberty. However, heterogeneity in children's individual level complicates challenge of personalized treatment. Our goal to quantify deviations from normative range morphometric variation children assess their associations clinical manifestations cognitive functions. We have developed sex-specific models on regional subcortical volume, cortical thickness, surface area data 457 healthy sourced two datasets. These were then utilized map (n = 187) sex- age-matched controls 108) another dataset. In group, exhibited a higher proportion regions infra-normal volumes, number correlated disease duration, seizure frequency, Raven's total score. findings suggest few extreme distributions heterogeneous are present minority individuals, emphasizing need monitor abnormalities throughout course disease.

Language: Английский

---

Atypical brain maturation in psychosis is associated with long-term cognitive decline and symptom progression

medRxiv (Cold Spring Harbor Laboratory), Journal Year: 2025, Volume and Issue: unknown

Published: April 4, 2025

ABSTRACT Background Clinical progression during psychosis has been closely associated with grey matter abnormalities resulting from atypical brain development. However, the complex interplay between psychopathology and neurodiversity challenges identifying neuroanatomical features that anticipate long-term cognitive symptomatic decline. Here, we collected MRI, cognitive, data 165 healthy controls 357 drug-naïve or minimally medicated FEP individuals were followed up 1,3,5 10 years after first episode. (1778 MRI scans assessments in total). Using normative modelling, derived subject-specific centile scores for cortical volume to investigate deviations their relationship deterioration. The association maps further characterized by examining cytoarchitectural neurobiological attributes using atlases. Aims To longitudinal exploring outcomes, as well underpinnings. Results centiles showed a widespread reduction at treatment initiation, analysis showing an increase time, indicating convergence toward normal maturation trajectories. Interestingly, this effect was reduced highly individuals. Additionally, found impairments experienced early stages correlated mitigated time. Positive symptomatology negatively regional centiles, higher benefited most treatment. Cytoarchitectural analyses revealed related FEP, function, specific molecular features, such serotonin receptor densities heteromodal areas. Conclusions Collectively, these findings underscore potential use of centile-based modelling better understanding how development contributes clinical neurodevelopmental conditions.

Language: Английский

---