Modern views of machine learning for precision psychiatry

Patterns, Journal Year: 2022, Volume and Issue: 3(11), P. 100602 - 100602

Published: Nov. 1, 2022

In light of the National Institute Mental Health (NIMH)'s Research Domain Criteria (RDoC), advent functional neuroimaging, novel technologies and methods provide new opportunities to develop precise personalized prognosis diagnosis mental disorders. Machine learning (ML) artificial intelligence (AI) are playing an increasingly critical role in era precision psychiatry. Combining ML/AI with neuromodulation can potentially explainable solutions clinical practice effective therapeutic treatment. Advanced wearable mobile also call for digital phenotyping health. this review, we a comprehensive review ML methodologies applications by combining neuromodulation, advanced psychiatry practice. We further molecular cross-species biomarker identification discuss AI (XAI) closed human-in-the-loop manner highlight potential multi-media information extraction multi-modal data fusion. Finally, conceptual practical challenges future research.

Language: Английский

---

Data-driven modelling of neurodegenerative disease progression: thinking outside the black box

Nature reviews. Neuroscience, Journal Year: 2024, Volume and Issue: 25(2), P. 111 - 130

Published: Jan. 8, 2024

Language: Английский

---

A multimodal machine learning system in early screening for toddlers with autism spectrum disorders based on the response to name

Frontiers in Psychiatry, Journal Year: 2023, Volume and Issue: 14

Published: Jan. 26, 2023

Reduced or absence of the response to name (RTN) has been widely reported as an early specific indicator for autism spectrum disorder (ASD), while few studies have quantified RTN toddlers with ASD in automatic way. The present study aims apply a multimodal machine learning system (MMLS) screening based on RTN.A total 125 were recruited, including (n = 61), developmental delay (DD, n 31), and typical (TD, 33). Procedures were, respectively, performed by evaluator caregiver. Behavioral data collected eight-definition tripod-mounted cameras coded MMLS. Response score, time, duration time accurately calculated evaluate RTN.Total accuracy scores rated computers was 0.92. In both caregiver procedures, had significant differences compared DD TD (all P-values < 0.05). area under curve (AUC) 0.81 computer-rated results, AUC 0.91 human-rated results. identification computer- results was, 74.8 82.9%. There difference between (Z 2.71, P-value 0.007).The can quantify behaviors procedures may effectively distinguish from non-ASD group. This novel provide low-cost approach identifying ASD. However, is not accurate human observer, detection single symptom like sufficient enough detect

Language: Английский

---

Exploring Digital Biomarkers of Illness Activity in Mood Episodes: Hypotheses Generating and Model Development Study

JMIR mhealth and uhealth, Journal Year: 2023, Volume and Issue: 11, P. e45405 - e45405

Published: March 20, 2023

Depressive and manic episodes within bipolar disorder (BD) major depressive (MDD) involve altered mood, sleep, activity, alongside physiological alterations wearables can capture. Firstly, we explored whether wearable data could predict (aim 1) the severity of an acute affective episode at intra-individual level 2) polarity euthymia among different individuals. Secondarily, which were related to prior predictions, generalization across patients, associations between symptoms data. We conducted a prospective exploratory observational study including patients with BD MDD on (manic, depressed, mixed) whose recorded using research-grade (Empatica E4) 3 consecutive time points (acute, response, remission episode). Euthymic healthy controls during single session (approximately 48 h). Manic assessed standardized psychometric scales. Physiological included following channels: acceleration (ACC), skin temperature, blood volume pulse, heart rate (HR), electrodermal activity (EDA). Invalid removed rule-based filter, channels aligned 1-second units segmented window lengths 32 seconds, as best-performing parameters. developed deep learning predictive models, channels' individual contribution permutation feature importance analysis, computed scales' items normalized mutual information (NMI). present novel, fully automated method for preprocessing analysis from device, viable supervised pipeline time-series analyses. Overall, 35 sessions (1512 hours) 12 mixed, euthymic) 7 (mean age 39.7, SD 12.6 years; 6/19, 32% female) analyzed. The mood was predicted moderate (62%-85%) accuracies 1), their (70%) accuracy 2). most relevant features former tasks ACC, EDA, HR. There fair agreement in classification (Kendall W=0.383). Generalization models unseen overall low accuracy, except models. ACC associated "increased motor activity" (NMI>0.55), "insomnia" (NMI=0.6), "motor inhibition" (NMI=0.75). EDA "aggressive behavior" (NMI=1.0) "psychic anxiety" (NMI=0.52). show potential identify specific mania depression quantitatively, both MDD. Motor stress-related (EDA HR) stand out digital biomarkers predicting depression, respectively. These findings represent promising pathway toward personalized psychiatry, allow early identification intervention episodes.

Language: Английский

---

Symptom dimensions of resting-state electroencephalographic functional connectivity in autism

Nature Mental Health, Journal Year: 2024, Volume and Issue: 2(3), P. 287 - 298

Published: Jan. 10, 2024

Language: Английский

---

Unravelling individual rhythmic abilities using machine learning

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: Jan. 11, 2024

Abstract Humans can easily extract the rhythm of a complex sound, like music, and move to its regular beat, in dance. These abilities are modulated by musical training vary significantly untrained individuals. The causes this variability multidimensional typically hard grasp single tasks. To date we lack comprehensive model capturing rhythmic fingerprints both musicians non-musicians. Here harnessed machine learning parsimonious abilities, based on behavioral testing (with perceptual motor tasks) individuals with without formal ( n = 79). We demonstrate that their link informal music experience be successfully captured profiles including minimal set measures. findings highlight techniques employed distill ultimately shed light individual relationship experiences.

Language: Английский

---

Rightward brain structural asymmetry in young children with autism

Molecular Psychiatry, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 15, 2025

Language: Английский

---

EEG-based major depressive disorder recognition by selecting discriminative features via stochastic search

Journal of Neural Engineering, Journal Year: 2023, Volume and Issue: 20(2), P. 026021 - 026021

Published: Feb. 22, 2023

Objective. Major depressive disorder (MDD) is a prevalent psychiatric whose diagnosis relies on experienced psychiatrists, resulting in low rate. As typical physiological signal, electroencephalography (EEG) has indicated strong association with human beings' mental activities and can be served as an objective biomarker for diagnosing MDD.Approach. The basic idea of the proposed method fully considers all channel information EEG-based MDD recognition designs stochastic search algorithm to select best discriminative features describing individual channels.Main results. To evaluate method, we conducted extensive experiments MODMA dataset (including dot-probe tasks resting state), 128-electrode public including 24 patients 29 healthy controls. Under leave-one-subject-out cross-validation protocol, achieved average accuracy 99.53% fear-neutral face pairs cued experiment 99.32% state, outperforming state-of-the-art methods. Moreover, our experimental results also that negative emotional stimuli could induce states, high-frequency EEG contributed significantly distinguishing between normal patients, which marker recognition.Significance. provided possible solution intelligent used develop computer-aided diagnostic tool aid clinicians early clinical purposes.

Language: Английский

---

Deep learning based joint fusion approach to exploit anatomical and functional brain information in autism spectrum disorders

Brain Informatics, Journal Year: 2024, Volume and Issue: 11(1)

Published: Jan. 9, 2024

Abstract Background: The integration of the information encoded in multiparametric MRI images can enhance performance machine-learning classifiers. In this study, we investigate whether combination structural and functional might improve performances a deep learning (DL) model trained to discriminate subjects with Autism Spectrum Disorders (ASD) respect typically developing controls (TD). Material methods We analyzed both brain scans publicly available within ABIDE I II data collections. considered 1383 male age between 5 40 years, including 680 ASD 703 TD from 35 different acquisition sites. extracted morphometric features Freesurfer CPAC analysis packages, respectively. Then, due multisite nature dataset, implemented harmonization protocol. vs. classification was carried out multiple-input DL model, consisting neural network which generates fixed-length feature representation each modality (FR-NN), Dense Neural Network for (C-NN). Specifically, joint fusion approach multiple source integration. main advantage latter is that loss propagated back FR-NN during training, thus creating informative representations modality. C-NN, number layers neurons per layer be optimized performs ASD-TD discrimination. evaluated by computing Area under Receiver Operating Characteristic curve nested 10-fold cross-validation. drive were identified SHAP explainability framework. Results AUC values 0.66±0.05 0.76±0.04 obtained discrimination when only or are considered, led an 0.78±0.04. set connectivity as most important two-class supports idea changes tend occur individuals regions belonging Default Mode Social Brain. Conclusions Our results demonstrate multimodal outperforms acquired single it efficiently exploits complementarity information.

Language: Английский

---

Discovering the gene-brain-behavior link in autism via generative machine learning

Science Advances, Journal Year: 2024, Volume and Issue: 10(24)

Published: June 12, 2024

Autism is traditionally diagnosed behaviorally but has a strong genetic basis. A genetics-first approach could transform understanding and treatment of autism. However, isolating the gene-brain-behavior relationship from confounding sources variability challenge. We demonstrate novel technique, 3D transport-based morphometry (TBM), to extract structural brain changes linked copy number variation (CNV) at 16p11.2 region. identified two distinct endophenotypes. In data Simons Variation in Individuals Project, detection these endophenotypes enabled 89 95% test accuracy predicting CNV images alone. Then, TBM direct visualization driving accurate prediction, revealing dose-dependent among deletion duplication carriers. These are sensitive articulation disorders explain portion intelligence quotient variability. Genetic stratification combined with reveal new many neurodevelopmental disorders, accelerating precision medicine, human neurodiversity.

Language: Английский

---