Machine learning techniques in a structural and functional MRI diagnostic approach in schizophrenia: a systematic review

目的 探讨精神分裂症患者脑静息态默认网络、中心执行网络及突显网络的网络内及网络间的功能连接特点.方法 对74例精神分裂症患者(患者组)和79名相匹配的健康者(对照组)进行静息态fMRI扫描.采取功能连接分析方法,以内侧前额叶皮质及右侧前岛叶为感兴趣区域分别提取默认网络、中心执行网络及突显网络,分析3个网络内及网络间静息态功能连接情况;并对有差异的功能连接与精神症状及病程等进行偏相关分析.结果 与对照组相比,患者组内侧前额叶皮质与双侧背外侧前额叶皮质(左侧-0.02±0.18与-0.12±0.12,t=4.25,P＜0.01;右侧-0.02±0.11与-0.10±0.10,t=5.40,P＜0.01)、双侧豆状壳核(左侧0.14±0.12与0.04±0.10,t=5.03,P＜0.01;右侧0.13±0.11与0.03±0.98,t=4.96,P＜0.01)的功能连接增强;右侧前岛叶与左侧额中回(-0.05±0.17与-0.07±0.17,t=4.60,P＜0.01)及楔前叶/后扣带回(0.01±0.13与-0.09±0.14,t=4.01,P＜0.01)的功能连接增强(均采用高斯随机场理论校正,体素水平P＜0.01,簇水平P＜0.05);默认网络与中心执行网络、突显网络间的负向功能连接减小甚至消失(Bonferroni校正P＜0.05).患者组PANSS阳性症状分与右侧前岛叶与楔前叶/后扣带回皮质间的功能连接呈正相关(r=0.232,P=0.025);患者组病程与右侧前岛叶与左侧额中回(r=0.288,P=0.007)、楔前叶/后扣带回(r=0.196,P=0.049)间的功能连接呈正相关.结论 精神分裂症患者存在静息态默认网络、中心执行网络及突显网络3个网络间的连接异常,可能与精神分裂症的病理机制有关.

Unprecedented 4D spatiotemporal infant regional cerebral blood flow framework and region-specific physiology–function coupling across infancy were elucidated, highlighting strong physiology–function coupling specifically at the default-mode network to meet extraneuronal metabolic demand for network emergence.

Cognition in ESRD may be improved by kidney transplantation, but mechanisms are unclear. We explored patterns of resting-state networks with resting-state functional magnetic resonance imaging among patients with ESRD before and after kidney transplantation. Thirty-seven patients with ESRD scheduled for kidney transplantation and 22 age-, sex-, and education-matched healthy subjects underwent resting-state functional magnetic resonance imaging. Patients were imaged before and 1 and 6 months after kidney transplantation. Functional connectivity of seven resting-state subnetworks was evaluated: default mode network, dorsal attention network, central executive network, self-referential network, sensorimotor network, visual network, and auditory network. Mixed effects models tested associations of ESRD, kidney transplantation, and neuropsychological measurements with functional connectivity. Compared with controls, pretransplant patients showed abnormal functional connectivity in six subnetworks. Compared with pretransplant patients, increased functional connectivity was observed in the default mode network, the dorsal attention network, the central executive network, the sensorimotor network, the auditory network, and the visual network 1 and 6 months after kidney transplantation (P=0.01). Six months after kidney transplantation, no significant difference in functional connectivity was observed for the dorsal attention network, the central executive network, the auditory network, or the visual network between patients and controls. Default mode network and sensorimotor network remained significantly different from those in controls when assessed 6 months after kidney transplantation. A relationship between functional connectivity and neuropsychological measurements was found in specific brain regions of some brain networks. The recovery patterns of resting-state subnetworks vary after kidney transplantation. The dorsal attention network, the central executive network, the auditory network, and the visual network recovered to normal levels, whereas the default mode network and the sensorimotor network did not recover completely 6 months after kidney transplantation. Neural resting-state functional connectivity was lower among patients with ESRD compared with control subjects, but it significantly improved with kidney transplantation. Resting-state subnetworks exhibited variable recovery, in some cases to levels that were no longer significantly different from those of normal controls.

With the rapid development of medicine and technology, machine learning (ML) techniques are extensively applied to medical informatics and the suboptimal health field to identify critical predictor variables and risk factors. Metabolic syndrome (MetS) and chronic kidney disease (CKD) are important risk factors for many comorbidities and complications. Existing studies that utilize different statistical or ML algorithms to perform CKD data analysis mostly analyze the early-stage subjects directly, but few studies have discussed the predictive models and important risk factors for the stage-III CKD high-risk health screening population. The middle stages 3a and 3b of CKD indicate moderate renal failure. This study aims to construct an effective hybrid important risk factor evaluation scheme for subjects with MetS and CKD stages III based on ML predictive models. The six well-known ML techniques, namely random forest (RF), logistic regression (LGR), multivariate adaptive regression splines (MARS), extreme gradient boosting (XGBoost), gradient boosting with categorical features support (CatBoost), and a light gradient boosting machine (LightGBM), were used in the proposed scheme. The data were sourced from the Taiwan health examination indicators and the questionnaire responses of 71,108 members between 2005 and 2017. In total, 375 stage 3a CKD and 50 CKD stage 3b CKD patients were enrolled, and 33 different variables were used to evaluate potential risk factors. Based on the results, the top five important variables, namely BUN, SBP, Right Intraocular Pressure (R-IOP), RBCs, and T-Cho/HDL-C (C/H), were identified as significant variables for evaluating the subjects with MetS and CKD stage 3a or 3b.

Background: Irritable Bowel Syndrome (IBS), as a functional gastrointestinal disorder, is characterized by abdominal pain and distension. Recent studies have shown that acupuncture treatment improves symptoms of diarrhea-predominant irritable bowel syndrome (IBS-D) by altering networks in certain brain regions. However, few studies have used resting-state functional magnetic resonance imaging (fMRI) to compare altered resting-state inter-network functional connectivity in IBS-D patients before and after acupuncture treatment.Objective: To analyze altered resting-state inter-network functional connectivity in IBS-D patients before and after acupuncture treatment.Methods: A total of 74 patients with IBS-D and 31 healthy controls (HCs) were recruited for this study. fMRI examination was performed in patients with IBS-D before and after acupuncture treatment, but only at baseline in HCs. Data on the left frontoparietal network (LFPN), default mode network (DMN), salience network (SN), ventral attention network (VAN), auditory network (AN), visual network (VN), sensorimotor network (SMN), dorsal attention network (DAN), and right frontoparietal network (RFPN) were subjected to independent component analysis (ICA). The functional connectivity values of inter-network were explored.Results: Acupuncture decreased irritable bowel syndrome symptom severity score (IBS-SSS) and Hamilton Anxiety Scale (HAMA). It also ameliorated symptoms related to IBS-D. Notably, functional connectivity between AN and VAN, SMN and DMN, RFPN and VAN in IBS-D patients after acupuncture treatment was different from that in HCs. Furthermore, there were differences in functional connectivity between DMN and DAN, DAN and LFPN, DMN and VAN before and after acupuncture treatment. The inter-network changes in DMN-VAN were positively correlated with changes in HAMA, life influence degree, and IBS-SSS in IBS-D.Conclusion: Altered inter-network functional connectivity is involved in several important hubs in large-scale networks. These networks are altered by acupuncture stimulation in patients with IBS-D.

BackgroundThree intrinsic connectivity networks in the brain, namely the central executive, salience, and default mode networks, have been identified as crucial to the understanding of higher cognitive functioning, and the functioning of these networks has been suggested to be impaired in psychopathology, including posttraumatic stress disorder (PTSD).Objective1) To describe three main large-scale networks of the human brain; 2) to discuss the functioning of these neural networks in PTSD and related symptoms; and 3) to offer hypotheses for neuroscientifically-informed interventions based on treating the abnormalities observed in these neural networks in PTSD and related disorders.MethodLiterature relevant to this commentary was reviewed.ResultsIncreasing evidence for altered functioning of the central executive, salience, and default mode networks in PTSD has been demonstrated. We suggest that each network is associated with specific clinical symptoms observed in PTSD, including cognitive dysfunction (central executive network), increased and decreased arousal/interoception (salience network), and an altered sense of self (default mode network). Specific testable neuroscientifically-informed treatments aimed to restore each of these neural networks and related clinical dysfunction are proposed.ConclusionsNeuroscientifically-informed treatment interventions will be essential to future research agendas aimed at targeting specific PTSD and related symptoms.

Schizophrenia is characterized by the distributed dysconnectivity of resting-state multiple brain networks. However, the abnormalities of intra- and inter-network functional connectivity (FC) in schizophrenia and its relationship to symptoms remain unknown. The aim of the present study is to compare the intra- and inter-connectivity of the intrinsic networks between a large sample of patients with schizophrenia and healthy controls. Using the Region of interest (ROI) to ROI FC analyses, the intra- and inter-network FC of the eight resting state networks [default mode network (DMN); salience network (SN); frontoparietal network (FPN); dorsal attention network (DAN); language network (LN); visual network (VN); sensorimotor network (SMN); and cerebellar network (CN)] were investigated in 196 schizophrenia and 169-healthy controls. Compared to the healthy control group, the schizophrenia group exhibited increased intra-network FC in the DMN and decreased intra-network FC in the CN. Additionally, the schizophrenia group showed the decreased inter-network FC mainly involved the SN-DMN, SN-LN and SN-CN while increased inter-network FC in the SN-SMN and SN-DAN (p < 0.05, FDR-corrected). Our study suggests widespread intra- and inter-network dysconnectivity among large-scale RSNs in schizophrenia, mainly involving the DMN, SN and SMN, which may further contribute to the dysconnectivity hypothesis of schizophrenia.

Frontoparietal areas link impairments of large-scale intrinsic brain networks with aberrant fronto-striatal interactions in OCD: a meta-analysis of resting-state functional connectivity

The aim of this study was to investigate the abnormities in functional connectivity (FC) within each modular network and between modular networks in patients with systemic lupus erythematosus (SLE). Twelve meaningful modular networks were identified via independent component analysis from 41 patients and 40 volunteers. Parametric tests were used to compare the intra- and intermodular FC between the groups. Partial correlation analysis was used to seek the relationships between abnormal FCs and the clinical data. Compared to the controls, SLE patients showed decreased intramodular FC in the anterior default mode network (aDMN), posterior default mode network (pDMN), ventral attention network (VAN), and sensorimotor network (SMN) and increased intramodular FC in the medial visual network (mVN) and left frontoparietal network. In addition, SLE patients showed decreased intermodular FC between the SMN and the lateral visual network (lVN), between the SMN and the VAN, and between the pDMN and the lVN and exhibited increased intermodular FC between the SMN and the salience network (SAN), between the pDMN and the SAN, and between the aDMN and the VAN. Moreover, we found several correlations among the abnormal FCs and the Mini-Mental State Examination in SLE patients. Mild cognitive impairment is compensated by the hyperconnectivity between the aDMN and the VAN, while severe cognitive impairment tends to be compensated by the hyperconnectivity between the SMN and the SAN. The FC value between the SMN and the SAN and between the aDMN and the VAN may serve as neuroimaging markers for monitoring cognitive progression in SLE patients.

Accumulating evidence suggests that hyper-connectivity is more characteristic of young children with autism spectrum disorder (ASD), while hypo-connectivity begins to emerge in adolescence and persists into adulthood. Despite increasing efforts being invested to explore the altered functional connectivity in ASD, the timing of the shift from intrinsic hyper-to hypo-connectivity of large-scale brain functional networks remains unclear. Here, we systematically depict the development of intrinsic functional connectivity in 800 participants from the Autism Brain Imaging Data Exchange. We first use independent component analyses to identify the large-scale brain functional networks. Then, we utilize the locally estimated scatterplot smoothing algorithm to fit the developmental trajectory of brain functional networks. Finally, we develop a "sliding threshold" method to detect the age stage at which the shift from hyper- to hypo-connectivity occurs in ASD. We identify six large-scale brain functional networks, including the default mode network (DMN), fronto-parietal network (FPN), salience network (SAN), auditory network (AN), somatomotor network (SMN), and visual network (VN). We find that primary networks (AN, SMN, and VN) undergo the shift from hyper- to hypo-connectivity earlier than high-order networks (DMN, FPN, and SAN) in ASD. At pre-adolescence, the SMN, AN, VN, DMN, SAN, and FPN undergo the shift from hyper- to hypo-connectivity in sequence in ASD. Our findings shed light on the age-related changes of intrinsic functional connectivity in ASD, highlighting the need for conceptualizing functional connectivity in ASD from a developmental perspective.

Fraud in the financial sector, e-commerce, and online services is becoming increasingly frequent and sophisticated. Traditional rule-based systems, while still helpful in detecting known fraud patterns, struggle to keep up with new, evolving attack vectors, as static rules are quickly circumvented. In contrast, machine learning (ML) provides a dynamic and scalable approach that can process vast amounts of transactional and behavioral data to identify subtle anomalies and suspicious activity.This paper provides a comprehensive overview of current ML techniques used in fraud detec-tion, categorized into three main groups: classification models, anomaly detection methods, and deep learning architectures. It discusses real-world applications across various fraud scenarios, including credit card abuse, account takeovers, cybercrime, and scams in digital comerce.Emphasis is placed on the strengths and limitations of each approach, with attention to real-world considerations like scalability, model transparency, and the challenge of class imbal-ance. The paper also reviews recent advances, including graph-based representations of fi-nancial interactions, IP-based behavioral profiling, and the emergence of hybrid systems that integrate multiple ML techniques – such as combining autoencoders with boosting algorithms for improved accuracy, especially when labeled data is scarce.The findings aim to support the development of flexible, high-performance fraud detection solutions that leverage the most effective ML practices and capitalize on the synergy of hybrid model architectures.

The aim of this paper is to review machine learning (ML) algorithms and techniques for hate speech detection in social media (SM). Hate speech problem is normally model as a text classification task. In this study, we examined the basic baseline components of hate speech classification using ML algorithms. There are five basic baseline components &#x2013; data collection and exploration, feature extraction, dimensionality reduction, classifier selection and training, and model evaluation, were reviewed. There have been improvements in ML algorithms that were employed for hate speech detection over time. New datasets and different performance metrics have been proposed in the literature. To keep the researchers informed regarding these trends in the automatic detection of hate speech, it calls for a comprehensive and an updated state-of-the-art. The contributions of this study are three-fold. First to equip the readers with the necessary information on the critical steps involved in hate speech detection using ML algorithms. Secondly, the weaknesses and strengths of each method is critically evaluated to guide researchers in the algorithm choice dilemma. Lastly, some research gaps and open challenges were identified. The different variants of ML techniques were reviewed which include classical ML, ensemble approach and deep learning methods. Researchers and professionals alike will benefit immensely from this study.

ObjectiveEnhanced technology in computer and internet has driven scale and quality of data to be improved in various areas including healthcare sectors. Machine Learning (ML) has played a pivotal role in efficiently analyzing those big data, but a general misunderstanding of ML algorithms still exists in applying them (e.g., ML techniques can settle a problem of small sample size, or deep learning is the ML algorithm). This paper reviewed the research of diagnosing mental illness using ML algorithm and suggests how ML techniques can be employed and worked in practice.MethodsResearches about mental illness diagnostic using ML techniques were carefully reviewed. Five traditional ML algorithms-Support Vector Machines (SVM), Gradient Boosting Machine (GBM), Random Forest, Naïve Bayes, and K-Nearest Neighborhood (KNN)-frequently used for mental health area researches were systematically organized and summarized.ResultsBased on literature review, it turned out that Support Vector Machines (SVM), Gradient Boosting Machine (GBM), Random Forest, Naïve Bayes, and K-Nearest Neighborhood (KNN) were frequently employed in mental health area, but many researchers did not clarify the reason for using their ML algorithm though every ML algorithm has its own advantages. In addition, there were several studies to apply ML algorithms without fully understanding the data characteristics.ConclusionResearchers using ML algorithms should be aware of the properties of their ML algorithms and the limitation of the results they obtained under restricted data conditions. This paper provides useful information of the properties and limitation of each ML algorithm in the practice of mental health.

Support vector machine classification and characterization of age-related reorganization of functional brain networks

Air quality, both indoors and outdoors, is crucial for public health as it affects respiratory conditions and overall well-being. Machine learning (ML) techniques offer innovative solutions for monitoring and predicting air quality. However, choosing the right ML algorithms and evaluation metrics is essential for creating accurate air quality prediction models, as these choices directly impact the accuracy and reliability of the results. This study aims to explore and summarize the literature on the use of ML techniques in predicting and optimizing air quality, addressing the urgent issue of air pollution. The review analyzed papers from two electronic databases, namely Scopus and Science Direct. Information on ML techniques for predicting air quality and identifying the main sources of pollution was extracted from 26 studies. The study focuses on common ML techniques employed in air quality prediction, including classification, deep learning, regression, ensemble learning, and combinations of regression and deep learning. It also identifies the evaluation metrics used to assess the performance of these models, such as recall, root mean square error (RMSE), mean absolute error (MAE), and coefficient of determination (R2). By synthesizing existing knowledge, this study provides a comprehensive understanding of the metrics researchers use to measure the overall effectiveness of ML algorithms. It serves as a benchmark for future research and guides the selection of appropriate evaluation metrics in the field of air quality prediction.