Multi-disease X-ray Image Classification of the Chest Based on Global and Local Fusion Adaptive Networks.

BACKGROUND: Chest X-ray image classification for multiple diseases is an important research direction in the field of computer vision and medical image processing. It aims to utilize advanced image processing techniques and deep learning algorithms to automatically analyze and identify X-ray images, determining whether specific pathologies or structural abnormalities exist in the images. OBJECTIVE: We present the MMPDenseNet network designed specifically for chest multi-label disease classification. METHODS: Initially, the network employs the adaptive activation function Meta-ACON to enhance feature representation. Subsequently, the network incorporates a multi-head self-attention mechanism, merging the conventional convolutional neural network with the Transformer, thereby bolstering the ability to extract both local and global features. Ultimately, the network integrates a pyramid squeeze attention module to capture spatial information and enrich the feature space. RESULTS: The concluding experiment yielded an average AUC of 0.898, marking an average accuracy improvement of 0.6% over the baseline model. When compared with the original network, the experimental results highlight that MMPDenseNet considerably elevates the classification accuracy of various chest diseases. CONCLUSION: It can be concluded that the network, thus, holds substantial value for clinical applications.

Multimodal frontal neuroimaging markers predict longitudinal craving reduction in abstinent individuals with heroin use disorder.

The variation in improvement among individuals with addiction after abstinence is a critical issue. Here, we aimed to identify robust multimodal markers associated with high response to 8-month abstinence in the individuals with heroin use disorder (HUD) and explore whether the identified markers could be generalized to the individuals with methamphetamine use disorder (MUD). According to the median of craving changes, 53 individuals with HUD with 8-month abstinence were divided into two groups: higher craving reduction and lower craving reduction. At baseline, clinical variables, cortical thickness and subcortical volume, fractional anisotropy (FA) of fibers and resting-state functional connectivity (RSFC) were extracted. Different strategies (single metric, multimodal neuroimaging fusion and multimodal neuroimaging-clinical data fusion) were used to identify reliable features for discriminating the individuals with HUD with higher craving reduction from those with lower reduction. The generalization ability of the identified features was validated in the 21 individuals with MUD. Multimodal neuroimaging-clinical fusion features with best performance was achieved an 87.1 ± 3.89% average accuracy in individuals with HUD, with a moderate accuracy of 66.7% when generalizing to individuals with MUD. The multimodal neuroimaging features, primarily converging in frontal regions (e.g., the left superior frontal (LSF) thickness, FA of the LSF-occipital tract, and RSFC of left middle frontal-right superior temporal lobe), collectively contributed to prediction alongside dosage and attention impulsiveness. In this study, we identified the validated multimodal frontal neuroimaging markers associated with higher response to long-term abstinence and revealed insights for the neural mechanisms of addiction abstinence, contributing to clinical strategies and treatment for addiction.

Altered morphometric similarity networks in insomnia disorder.

Previous studies on structural covariance network (SCN) suggested that patients with insomnia disorder (ID) show abnormal structural connectivity, primarily affecting the somatomotor network (SMN) and default mode network (DMN). However, evaluating a single structural index in SCN can only reveal direct covariance relationship between two brain regions, failing to uncover synergistic changes in multiple structural features. To cover this research gap, the present study utilized novel morphometric similarity networks (MSN) to examine the morphometric similarity between cortical areas in terms of multiple sMRI parameters measured at each area. With seven T1-weighted imaging morphometric features from the Desikan-Killiany atlas, individual MSN was constructed for patients with ID (N = 87) and healthy control groups (HCs, N = 84). Two-sample t-test revealed differences in MSN between patients with ID and HCs. Correlation analyses examined associations between MSNs and sleep quality, insomnia symptom severity, and depressive symptoms severity in patients with ID. The right paracentral lobule (PCL) exhibited decreased morphometric similarity in patients with ID compared to HCs, mainly manifested by its de-differentiation (meaning loss of distinctiveness) with the SMN, DMN, and ventral attention network (VAN), as well as its decoupling with the visual network (VN). Greater PCL-based de-differentiation correlated with less severe insomnia and fewer depressive symptoms in the patients group. Additionally, patients with less depressive symptoms showed greater PCL de-differentiation from the SMN. As an important pilot step in revealing the underlying morphometric similarity alterations in insomnia disorder, the present study identified the right PCL as a hub region that is de-differentiated with other high-order networks. Our study also revealed that MSN has an important potential to capture clinical significance related to insomnia disorder.

Electroencephalography microstates as novel functional biomarkers for insomnia disorder.

Background: Insomnia disorder (ID) is one of the most common mental disorders. Research on ID focuses on exploring its mechanism of disease, novel treatments and treatment outcome prediction. An emerging technique in this field is the use of electroencephalography (EEG) microstates, which offer a new method of EEG feature extraction that incorporates information from both temporal and spatial dimensions. Aims: To explore the electrophysiological mechanisms of repetitive transcranial magnetic stimulation (rTMS) for ID treatment and use baseline microstate metrics for the prediction of its efficacy. Methods: This study included 60 patients with ID and 40 age-matched and gender-matched good sleep controls (GSC). Their resting-state EEG microstates were analysed, and the Pittsburgh Sleep Quality Index (PSQI) and polysomnography (PSG) were collected to assess sleep quality. The 60 patients with ID were equally divided into active and sham groups to receive rTMS for 20 days to test whether rTMS had a moderating effect on abnormal microstates in patients with ID. Furthermore, in an independent group of 90 patients with ID who received rTMS treatment, patients were divided into optimal and suboptimal groups based on their median PSQI reduction rate. Baseline EEG microstates were used to build a machine-learning predictive model for the effects of rTMS treatment. Results: The class D microstate was less frequent and contribute in patients with ID, and these abnormalities were associated with sleep onset latency as measured by PSG. Additionally, the abnormalities were partially reversed to the levels observed in the GSC group following rTMS treatment. The baseline microstate characteristics could predict the therapeutic effect of ID after 20 days of rTMS, with an accuracy of 80.13%. Conclusions: Our study highlights the value of EEG microstates as functional biomarkers of ID and provides a new perspective for studying the neurophysiological mechanisms of ID. In addition, we predicted the therapeutic effect of rTMS on ID based on the baseline microstates of patients with ID. This finding carries great practical significance for the selection of therapeutic options for patients with ID.

Covariance patterns between sleep health domains and distributed intrinsic functional connectivity.

Sleep health is both conceptually and operationally a composite concept containing multiple domains of sleep. In line with this, high dependence and interaction across different domains of sleep health encourage a transition in sleep health research from categorical to dimensional approaches that integrate neuroscience and sleep health. Here, we seek to identify the covariance patterns between multiple sleep health domains and distributed intrinsic functional connectivity by applying a multivariate approach (partial least squares). This multivariate analysis reveals a composite sleep health dimension co-varying with connectivity patterns involving the attentional and thalamic networks and which appear relevant at the neuromolecular level. These findings are further replicated and generalized to several unseen independent datasets. Critically, the identified sleep-health related connectome shows diagnostic potential for insomnia disorder. These results together delineate a potential brain connectome biomarker for sleep health with high potential for clinical translation.

The changes of intrinsic connectivity contrast in young smokers.

Previous studies demonstrated that reward circuit plays an important role in smoking. The differences of functional and structural connectivity were found among several brain regions such as thalamus and frontal lobe. However, few studies focused on functional connectivity (FC) in whole-brain voxel level of young smokers. In this study, intrinsic connectivity contrast (ICC) was used to perform voxel-based whole-brain analyses in 55 young smokers and 55 matched non-smokers to identify brain regions with significant group differences. ICC results showed that the connectivity of young smokers in medial frontal cortex (MedFC), supramarginal gyrus anterior division left (L_aSMG), central opercular cortex left (L_CO) and middle frontal gyrus left (L_MidFG) showed a significantly lower trend compared with the non-smokers. The seed-based FC analysis about MedFC indicated that young smokers showed reduced connectivity between the MedFC and left hippocampus, left amygdala compared to non-smokers. Correlation analysis showed that the ICC of MedFC in young smokers was significantly negatively correlated with Fagerstrom test for nicotine dependence (FTND) and Questionnaire on Smoking Urges (QSU). The FC between the MedFC and left hippocampus, left amygdala was significantly negatively correlated with Pack_years. The mediation analysis indicated that ICC of MedFC completely mediated FTND and QSU of young smokers. The results suggest that nicotine accumulation may affect the communication of the frontal lobe with the whole brain to some extent, leading to changes in smoking cravings. The above research also provides in-depth insights into the mechanism of adolescent smoking addiction and related intervention treatment.

Glutamate levels in the ventromedial prefrontal cortex and resting-state functional connectivity within reward circuits in alcohol-dependent patients.

Great progress has been made in understanding the neural mechanisms associated with alcohol-dependent (AD) patients. However, the interactions within the reward circuits of the patients need further exploration. Glutamatergic projections from the prefrontal cortex to some brain regions are present in the reward circuit. However, little is known about the potential implications of glutamate levels in the prefrontal cortex on abnormal interactions within reward circuits in AD patients. To determine the potential roles of reward circuits in drinking, we investigated differences in resting-state functional connectivity (RSFC) and multivariate Granger causality analysis between 20 AD patients and 20 healthy controls (HC). The neuroimaging findings were then correlated with clinical variables (alcohol use disorder identification test). The ventromedial prefrontal cortex (VmPFC) is believed to play a critical role in addiction disorders, and glutamatergic projections from the prefrontal cortex to several regions of the brain are present in reward circuits. Proton magnetic resonance spectroscopy was also performed to assess the difference in glutamate levels in VmPFC between AD patients and HC. The results showed that the strength of functional connectivity in the reward circuit was generally attenuated in AD patients, and the reciprocal enhancement of activity between the right insula, left thalamus and VmPFC was found to be significantly greater in AD patients. It is worth noting that although glutamate levels in the VmPFC did not show significant differences between the two groups, the level of glutamate in the VmPFC was significantly correlated with RSFC. We hope that the current findings will help us to develop new intervention models based on the important role of the VmPFC in AD.

Effects on resting-state EEG phase-amplitude coupling in insomnia disorder patients following 1 Hz left dorsolateral prefrontal cortex rTMS.

Despite burgeoning evidence for cortical hyperarousal in insomnia disorder, the existing results on electroencephalography spectral features are highly heterogeneous. Phase-amplitude coupling, which refers to the modulation of the low-frequency phase to a high-frequency amplitude, is probably a more sensitive quantitative measure for characterizing abnormal neural oscillations and explaining the therapeutic effect of repetitive transcranial magnetic stimulation in the treatment of patients with insomnia disorder. Sixty insomnia disorder patients were randomly divided into the active and sham treatment groups to receive 4 weeks of repetitive transcranial magnetic stimulation treatment. Behavioral assessments, resting-state electroencephalography recordings, and sleep polysomnography recordings were performed before and after repetitive transcranial magnetic stimulation treatment. Forty good sleeper controls underwent the same assessment. We demonstrated that phase-amplitude coupling values in the frontal and temporal lobes were weaker in Insomnia disorder patients than in those with good sleeper controls at baseline and that phase-amplitude coupling values near the intervention area were significantly enhanced after active repetitive transcranial magnetic stimulation treatment. Furthermore, the enhancement of phase-amplitude coupling values was significantly correlated with the improvement of sleep quality. This study revealed the potential of phase-amplitude coupling in assessing the severity of insomnia disorder and the efficacy of repetitive transcranial magnetic stimulation treatment, providing new insights on the abnormal physiological mechanisms and future treatments for insomnia disorder.

The potential of electroencephalography coherence to predict the outcome of repetitive transcranial magnetic stimulation in insomnia disorder.

BACKGROUND: It is unknown whether repetitive Transcranial Magnetic Stimulation (rTMS) could improve sleep quality by modulating electroencephalography (EEG) connectivity of insomnia disorder (ID) patients. Great heterogeneity had been found in the clinical outcomes of rTMS for ID. The study aimed to investigate the potential mechanisms of rTMS therapy for ID and develop models to predict clinical outcomes. METHODS: In Study 1, 50 ID patients were randomly divided into active and sham groups, and subjected to 20 sessions of treatment with 1 Hz rTMS over the left dorsolateral prefrontal cortex. EEG during awake, Polysomnography, and clinical assessment were collected and analyzed before and after rTMS. In Study 2, 120 ID patients were subjected to active rTMS stimulation and were then separated into optimal and sub-optimal groups due to the median of Pittsburgh Sleep Quality Index reduction rate. Machine learning models were developed based on baseline EEG coherence to predict rTMS treatment effects. RESULTS: In Study 1, decreased EEG coherence in theta and alpha bands were observed after rTMS treatment, and changes in theta band (F7-O1) coherence were correlated with changes in sleep efficiency. In Study 2, baseline EEG coherence in theta, alpha, and beta bands showed the potential to predict the treatment effects of rTMS for ID. CONCLUSION: rTMS improved sleep quality of ID patients by modulating the abnormal EEG coherence. Baseline EEG coherence between certain channels in theta, alpha, and beta bands could act as potential biomarkers to predict the therapeutic effects.

Reconfigurations of Dynamic Functional Network Connectivity in Large-scale Brain Network after Prolonged Abstinence in Heroin Users.

BACKGROUND: Brain recovery phenomenon after long-term abstinence had been reported in substance use disorders. Yet, few longitudinal studies have been conducted to observe the abnormal dynamic functional connectivity (dFNC) of large-scale brain networks and recovery after prolonged abstinence in heroin users. OBJECTIVE: The current study will explore the brain network dynamic connection reconfigurations after prolonged abstinence in heroin users (HUs). METHODS: The 10-month longitudinal design was carried out for 40 HUs. The 40 healthy controls (HCs) were also enrolled. Group independent component analysis (GICA) and dFNC analysis were employed to detect the different dFNC patterns of addiction-related ICNs between HUs and HCs. The temporal properties and the graph-theoretical properties were calculated. Whether the abnormalities would be reconfigured in HUs after prolonged abstinence was then investigated. RESULTS: Based on eight functional networks extracted from GICA, four states were identified by the dFNC analysis. Lower mean dwell time and fraction rate in state4 were found for HUs, which were increased toward HCs after prolonged abstinence. In this state, HUs at baseline showed higher dFNC of RECN-aSN, aSN- aSN and dDMN-pSN, which decreased after protracted abstinence. A similar recovery phenomenon was found for the global efficiency and path length in abstinence HUs. Mean while, the abnormal dFNC strength was correlated with craving both at baseline and after abstinence. CONCLUSION: Our longitudinal study observed the large-scale brain network reconfiguration from the dynamic perspective in HUs after prolonged abstinence and improved the understanding of the neurobiology of prolonged abstinence in HUs.

Increased resting-state alpha coherence and impaired inhibition control in young smokers.

Exposure to nicotine is the first cause of entirely preventable death killing, which is commonly initiated in adolescence. Previous studies revealed the changes of electroencephalography (EEG) and inhibition control in smokers. However, little is known about the specific link between alpha coherence during the resting state and inhibition control ability in young smokers. The present study aimed to investigate inter-hemispherical and frontal-parietal alpha coherence changes and assessed the relationships between alpha coherence and inhibition control in young smokers. We collected resting-state EEG data from 23 young smokers and 24 healthy controls. Inhibition control ability was assessed by a Go/NoGo task. Compared to healthy controls, young smokers exhibited increased inter-hemispherical and frontal-parietal alpha coherence. Furthermore, young smokers committed more NoGo errors in the Go/NogGo task. It is noteworthy that alpha coherence at the frontal electrode sites was positively correlated with NoGo errors in healthy controls, whereas inverse correlations were observed in young smokers. Our findings suggested that alterations of alpha coherence may provide support to the earlier nicotine-dependence-related research findings, which may help us to understand the neuropathology of inhibitory control in young smokers.

Altered resting-state electroencephalography microstate characteristics in young male smokers.

The development of nicotine addiction was associated with the abnormalities of intrinsic functional networks during the resting state in young adult smokers. As a whole-brain imaging approach, EEG microstate analysis treated multichannel EEG recordings as a series of quasi-steady microscopic states which were related to the resting-state networks (RSNs) found by fMRI. The aim of this study was to examine whether the resting-state EEG microstate analysis may provide novel insights into the abnormal temporal properties of intrinsic brain activities in young smokers. We used 64-channel resting-state EEG datasets to investigate alterations in microstate characteristics between twenty-five young smokers and 25 age- and gender-matched non-smoking controls. Four classic EEG microstates (microstate A, B, C, and D) were obtained, and the four temporal parameters of each microstate were extracted, i.e., duration, occurrence, coverage, and transition probabilities. Compared with non-smoking controls, young smokers showed decreased occurrence of microstate C and increased duration of microstate D. Furthermore, both the duration and coverage of microstate D were significantly negatively correlated with Fagerstrom Test of Nicotine Dependence (FTND) in young smoker group. The complex changes in the microstate time-domain parameters might correspond to the abnormalities of RSNs in analyses of FC measured with fMRI in the previous studies and indicate the altered specific brain functions in young smokers. Microstate D could be potentially represented as a selective biomarker for predicting the dependence degree of adolescent smokers on cigarettes. These results suggested that EEG microstate analysis might detect the deviant functions of large-scale cortical activities in young smokers and provide a new perspective for the study of brain networks of adolescent smokers.

[Application and challenges of neuroimaging technology in mechanisms and treatment of drug addiction].

Neuroimaging technologies can non-invasively characterize the structure and function of addiction brain, reveal the neural mechanism of addictive behavior, and provide a priori for the potential targets of brain stimulation. Neuroimaging technologies has played an important role in the study of drug addiction diseases, relapse prediction, and therapeutic evaluation of non-invasive brain stimulations, but it also faces many challenges. In this manuscript, we discuss the classification and analysis methods of neuroimaging technologies, its application in addiction and challenges, thus to promote the application of neuroimaging technology in the treatment of drug addiction.

Potential brain recovery of frontostriatal circuits in heroin users after prolonged abstinence: A preliminary study.

Neuroscientists have devoted efforts to explore potential brain recovery after prolonged abstinence in heroin users (HU). However, not much is known about whether frontostriatal circuits can recover after prolonged abstinence in HU. An eight-month longitudinal study was carried out for HU. Two MRI scans were obtained at baseline (HU1) and 8-month follow-up (HU2). The functional and structural connectivities of dorsal and ventral frontostriatal pathways were measured by resting-state functional connectivity (RSFC) and diffusion tensor imaging (DTI). Correlation analyses were employed to reveal the associations between neuroimaging and behavioral changes. Results suggested that relative to healthy controls (HCs), HU1 showed lower fractional anisotropy (FA) in the right dorsolateral prefrontal cortex (DLPFC)-to-caudate tracts and medial orbitofrontal cortex (mOFC)-to-nucleus accumbens (NAc) tracts as well as decreased RSFC in the left mOFC-NAc circuits. Longitudinal results revealed reduced craving and enhanced cognitive control in HU2 compared with HU1. After prolonged abstinence, HU2 showed increased FA values in the right DLPFC-caudate and mOFC-NAc tracts as well as increased RSFC strength in the bilateral mOFC-NAc circuits compared with HU1. In addition, changes in RSFC and FA values in the right mOFC-NAc circuit were negatively correlated with craving score changes. Similarly, negative correlations were also found between changes of RSFC in the bilateral DLPFC-caudate circuits and TMT-A scores. We provided scientific evidence for brain recovery of the dorsal and ventral frontostriatal circuits in HU after prolonged abstinence, and these circuits may be potential neuroimaging biomarkers for cognition and craving changes.

Right arcuate fasciculus and left uncinate fasciculus abnormalities in young smoker.

Previous diffusion tensor imaging (DTI) studies had investigated the white matter (WM) integrity abnormalities in smokers. Exposure to nicotine disrupts neurodevelopment during adolescence, possibly by disrupting the trophic effects of acetylcholine. However, little is known about the diffusion parameters of specific fibre bundles at multiple locations in young smokers. Thirty-seven young smokers and 29 age-, education- and gender-matched healthy non-smokers participated in this study. Automated Fibre Quantification (AFQ) was employed to investigate the WM microstructure in young smokers by integrating multiple indices. Diffusion parameters, that is, fractional anisotropy (FA), axial diffusion (AD), radial diffusion (RD) and mean diffusion (MD), were calculated at 100 points along the length of 18 major brain tracts. The relationships between neuroimaging differences and smoking behaviours were explored, including Fagerström Test of Nicotine Dependence (FTND) and pack-years. Compared with non-smokers, young smokers showed significantly increased FA, AD and decreased RD in the left uncinate fasciculus (UF) and right thalamic radiation (TR), increased AD, RD and decreased FA in the right arcuate fasciculus (Arc). Correlation analyses revealed that FA values of the left UF and RD values of the right Arc were negatively correlated with FTND score in smokers and FA values of the right Arc were positively correlated with FTND scores. Positive correlation was observed between AD values of the left UF and pack-years in smokers. The findings enhanced our understanding of the potential effect of adolescent smoking on WM microstructure.