Innovative Neuroimaging Biomarker Distinction of Major Depressive Disorder and Bipolar Disorder through Structural Connectome Analysis and Machine Learning Models.

Major depressive disorder (MDD) and bipolar disorder (BD) share clinical features, which complicates their differentiation in clinical settings. This study proposes an innovative approach that integrates structural connectome analysis with machine learning models to discern individuals with MDD from individuals with BD. High-resolution MRI images were obtained from individuals diagnosed with MDD or BD and from HCs. Structural connectomes were constructed to represent the complex interplay of brain regions using advanced graph theory techniques. Machine learning models were employed to discern unique connectivity patterns associated with MDD and BD. At the global level, both BD and MDD patients exhibited increased small-worldness compared to the HC group. At the nodal level, patients with BD and MDD showed common differences in nodal parameters primarily in the right amygdala and the right parahippocampal gyrus when compared with HCs. Distinctive differences were found mainly in prefrontal regions for BD, whereas MDD was characterized by abnormalities in the left thalamus and default mode network. Additionally, the BD group demonstrated altered nodal parameters predominantly in the fronto-limbic network when compared with the MDD group. Moreover, the application of machine learning models utilizing structural brain parameters demonstrated an impressive 90.3% accuracy in distinguishing individuals with BD from individuals with MDD. These findings demonstrate that combined structural connectome and machine learning enhance diagnostic accuracy and may contribute valuable insights to the understanding of the distinctive neurobiological signatures of these psychiatric disorders.

Aberrant brain network topology in youth with a familial risk for bipolar disorder: a task-based fMRI connectome study.

BACKGROUND: Youth with a family history of bipolar disorder (BD) may be at increased risk for mood disorders and for developing side effects after antidepressant exposure. The neurobiological basis of these risks remains poorly understood. We aimed to identify biomarkers underlying risk by characterizing abnormalities in the brain connectome of symptomatic youth at familial risk for BD. METHODS: Depressed and/or anxious youth (n = 119, age = 14.9 ± 1.6 years) with a family history of BD but no prior antidepressant exposure and typically developing controls (n = 57, age = 14.8 ± 1.7 years) received functional magnetic resonance imaging (fMRI) during an emotional continuous performance task. A generalized psychophysiological interaction (gPPI) analysis was performed to compare their brain connectome patterns, followed by machine learning of topological metrics. RESULTS: High-risk youth showed weaker connectivity patterns that were mainly located in the default mode network (DMN) (network weight = 50.1%) relative to controls, and connectivity patterns derived from the visual network (VN) constituted the largest proportion of aberrant stronger pairs (network weight = 54.9%). Global local efficiency (Elocal, p = .022) and clustering coefficient (Cp, p = .029) and nodal metrics of the right superior frontal gyrus (SFG) (Elocal: p < .001; Cp: p = .001) in the high-risk group were significantly higher than those in healthy subjects, and similar patterns were also found in the left insula (degree: p = .004; betweenness: p = .005; age-by-group interaction, p = .038) and right hippocampus (degree: p = .003; betweenness: p = .003). The case-control classifier achieved a cross-validation accuracy of 78.4%. CONCLUSIONS: Our findings of abnormal connectome organization in the DMN and VN may advance mechanistic understanding of risk for BD. Neuroimaging biomarkers of increased network segregation in the SFG and altered topological centrality in the insula and hippocampus in broader limbic systems may be used to target interventions tailored to mitigate the underlying risk of brain abnormalities in these at-risk youth.

Differential cortical gray matter changes in early- and late-onset patients with amyotrophic lateral sclerosis.

Age at onset may be an important feature associated with distinct subtypes of amyotrophic lateral sclerosis (ALS). Little is known about the neuropathological mechanism of early-onset ALS (EO-ALS) and late-onset ALS (LO-ALS). Ninety ALS patients were divided into EO-ALS and LO-ALS group, and 128 healthy controls were matched into young controls(YCs) and old controls (OCs). A voxel-based morphometry approach was employed to investigate differences in gray matter volume (GMV). Significant age at onset-by-diagnosis interactions were found in the left parietal operculum, left precentral gyrus, bilateral postcentral gyrus, right occipital gyrus, and right orbitofrontal cortex. Post hoc analysis revealed a significant decrease in GMV in all affected regions of EO-ALS patients compared with YCs, with increased GMV in 5 of the 6 brain regions, except for the right orbitofrontal cortex, in LO-ALS patients compared with OCs. LO-ALS patients had a significantly increased GMV than EO-ALS patients after removing the aging effect. Correspondingly, GMV of the left postcentral gyrus correlated with disease severity in the 2 ALS groups. Our findings suggested that the pathological mechanisms in ALS patients with different ages at onset might differ. These findings provide unique insight into the clinical and biological heterogeneity of the 2 ALS subtypes.

Effects of PRRT2 mutation on brain gray matter networks in paroxysmal kinesigenic dyskinesia.

Although proline-rich transmembrane protein 2 is the primary causative gene of paroxysmal kinesigenic dyskinesia, its effects on the brain structure of paroxysmal kinesigenic dyskinesia patients are not yet clear. Here, we explored the influence of proline-rich transmembrane protein 2 mutations on similarity-based gray matter morphological networks in individuals with paroxysmal kinesigenic dyskinesia. A total of 51 paroxysmal kinesigenic dyskinesia patients possessing proline-rich transmembrane protein 2 mutations, 55 paroxysmal kinesigenic dyskinesia patients possessing proline-rich transmembrane protein 2 non-mutation, and 80 healthy controls participated in the study. We analyzed the structural connectome characteristics across groups by graph theory approaches. Relative to paroxysmal kinesigenic dyskinesia patients possessing proline-rich transmembrane protein 2 non-mutation and healthy controls, paroxysmal kinesigenic dyskinesia patients possessing proline-rich transmembrane protein 2 mutations exhibited a notable increase in characteristic path length and a reduction in both global and local efficiency. Relative to healthy controls, both patient groups showed reduced nodal metrics in right postcentral gyrus, right angular, and bilateral thalamus; Relative to healthy controls and paroxysmal kinesigenic dyskinesia patients possessing proline-rich transmembrane protein 2 non-mutation, paroxysmal kinesigenic dyskinesia patients possessing proline-rich transmembrane protein 2 mutations showed almost all reduced nodal centralities and structural connections in cortico-basal ganglia-thalamo-cortical circuit including bilateral supplementary motor area, bilateral pallidum, and right caudate nucleus. Finally, we used support vector machine by gray matter network matrices to classify paroxysmal kinesigenic dyskinesia patients possessing proline-rich transmembrane protein 2 mutations and paroxysmal kinesigenic dyskinesia patients possessing proline-rich transmembrane protein 2 non-mutation, achieving an accuracy of 73%. These results show that proline-rich transmembrane protein 2 related gray matter network deficits may contribute to paroxysmal kinesigenic dyskinesia, offering new insights into its pathophysiological mechanisms.

Common and distinct neural correlates of emotional processing in individuals at familial risk for major depressive disorder and bipolar disorder: A comparative meta-analysis.

Individuals at familial risk for mood disorders exhibit deficits in emotional processing and associated brain dysfunction prior to illness onset. However, such brain-behavior abnormalities related to familial predisposition remain poorly understood. To investigate robust abnormal functional activation patterns during emotional processing in unaffected at-risk relatives of patients with major depressive disorder (UAR-MDD) and bipolar disorder (UAR-BD), we performed a meta-analysis of task-based functional magnetic resonance imaging studies using Seed-based d Mapping (SDM) toolbox. Common and distinct patterns of abnormal functional activation between UAR-MDD and UAR-BD were detected via conjunction and differential analyses. A total of 17 studies comparing 481 UAR and 670 healthy controls (HC) were included. Compared with HC, UAR-MDD exhibited hyperactivation in the parahippocampal gyrus, amygdala and cerebellum, while UAR-BD exhibited parahippocampal hyperactivation and hypoactivation in the striatum and middle occipital gyrus (MOG). Conjunction analysis revealed shared hyperactivated PHG in both groups. Differential analysis indicated that the activation patterns of amygdala and MOG significantly differed between UAR-MDD and UAR-BD. These findings provide novel insights into common and distinct neural phenotypes for familial risk and associated risk mechanisms in MDD and BD, which may have implications in guiding precise prevention strategies tailored to the family context.

Cortical and subcortical structural differences in psychostimulant-free ADHD youth with and without a family history of bipolar I disorder: a cross-sectional morphometric comparison.

Although attention-deficit/hyperactivity disorder (ADHD) and a family history of bipolar I disorder (BD) are associated with increased risk for developing BD, their neuroanatomical substrates remain poorly understood. This study compared cortical and subcortical gray matter morphology in psychostimulant-free ADHD youth with and without a first-degree relative with BD and typically developing healthy controls. ADHD youth (ages 10-18 years) with ('high-risk', HR) or without ('low-risk', LR) a first-degree relative with BD and healthy comparison youth (HC) were enrolled. High-resolution 3D T1-weighted images were acquired using a Philips 3.0 T MR scanner. The FreeSurfer image analysis suite was used to measure cortical thickness, surface area, and subcortical volumes. A general linear model evaluated group differences in MRI features with age and sex as covariates, and exploratory correlational analyses evaluated associations with symptom ratings. A total of n = 142 youth (mean age: 14.16 ± 2.54 years, 35.9% female) were included in the analysis (HC, n = 48; LR, n = 49; HR, n = 45). The HR group exhibited a more severe symptom profile, including higher mania and dysregulation scores, compared to the LR group. For subcortical volumes, the HR group exhibited smaller bilateral thalamic, hippocampal, and left caudate nucleus volumes compared to both LR and HC, and smaller right caudate nucleus compared with LR. No differences were found between LR and HC groups. For cortical surface area, the HR group exhibited lower parietal and temporal surface area compared with HC and LR, and lower orbitofrontal and superior frontal surface area compared to LR. The HR group exhibited lower left anterior cingulate surface area compared with HC. LR participants exhibited greater right pars opercularis surface area compared with the HC. Some cortical alterations correlated with symptom severity ratings. These findings suggest that ADHD in youth with a BD family history is associated with a more a severe symptom profile and a neuroanatomical phenotype that distinguishes it from ADHD without a BD family history.

Dynamic alterations of striatal-related functional networks in juvenile absence epilepsy.

PURPOSE: To explore the features of dynamic functional connectivity (dFC) variability of striatal-cortical/subcortical networks in juvenile absence epilepsy (JAE). METHODS: We collected resting-state functional magnetic imaging data from 18 JAE patients and 28 healthy controls. The striatum was divided into six pairs of regions: the inferior-ventral striatum (VSi), superior-ventral striatum (VSs), dorsal-caudal putamen, dorsal-rostral putamen, dorsal-caudate (DC) and ventral-rostral putamen. We assessed the dFC variability of each subdivision in the whole brain using the sliding-window method, and correlated altered circuit with clinical variables in JAE patients. RESULTS: We found altered dFC variability of striatal-cortical/subcortical networks in patients with JAE. The VSs exhibited decreased dFC variability with subcortical regions, and dFC variability between VSs and thalamus was negatively correlated with epilepsy duration. For the striatal-cortical networks, the dFC variability was decreased in VSi-affective network but increased in DC-executive network. The altered dynamics of striatal-cortical networks involved crucial nodes of the default mode network (DMN). CONCLUSION: JAE patients exhibit excessive stability in the striatal-subcortical networks. For striatal-cortical networks in JAE, the striatal-affective circuit was more stable, while the striatal-executive circuit was more variable. Furthermore, crucial nodes of DMN were changed in striatal-cortical networks in JAE.

Brain network structural connectome abnormalities among youth with attention-deficit/hyperactivity disorder at varying risk for bipolar I disorder: a cross-sectional graph-based magnetic resonance imaging study.

BACKGROUND: Attention-deficit/hyperactivity disorder (ADHD) is highly prevalent among youth with or at familial risk for bipolar-I disorder (BD-I), and ADHD symptoms commonly precede and may increase the risk for BD-I; however, associated neuropathophysiological mechanisms are not known. In this cross-sectional study, we sought to investigate brain structural network topology among youth with ADHD, with and without familial risk of BD-I. METHODS: We recruited 3 groups of psychostimulant-free youth (aged 10-18 yr), namely youth with ADHD and at least 1 biological parent or sibling with BD-I (high-risk group), youth with ADHD who did not have a first- or second-degree relative with a mood or psychotic disorder (low-risk group) and healthy controls. We used graph-based network analysis of structural magnetic resonance imaging data to investigate topological properties of brain networks. We also evaluated relationships between topological metrics and mood and ADHD symptom ratings. RESULTS: A total of 149 youth were included in the analysis (49 healthy controls, 50 low-risk youth, 50 high-risk youth). Low-risk and high-risk ADHD groups exhibited similar differences from healthy controls, mainly in the default mode network and central executive network. We found topological alterations in the salience network of the high-risk group, relative to both low-risk and control groups. We found significant abnormalities in global network properties in the high-risk group only, compared with healthy controls. Among both low-risk and high-risk ADHD groups, nodal metrics in the right triangular inferior frontal gyrus correlated positively with ADHD total and hyperactivity/impulsivity subscale scores. LIMITATIONS: The cross-sectional design of this study could not determine the relevance of these findings to BD-I risk progression. CONCLUSION: Youth with ADHD, with and without familial risk for BD-I, exhibit common regional abnormalities in the brain connectome compared with healthy youth, whereas alterations in the salience network distinguish these groups and may represent a prodromal feature relevant to BD-I risk.

More extensive structural damage in temporal lobe epilepsy with hippocampal sclerosis type 1.

OBJECTIVE: To explore clinical and structural differences between mesial temporal lobe epilepsy (mTLE) patients with different hippocampal sclerosis (HS) subtypes. METHODS: High-resolution T1-weighted MRI and diffusion tensor imaging data were obtained in 41 refractory mTLE patients and 52 age- and sex-matched healthy controls. Postoperative histopathological examination confirmed HS type 1 in 30 patients and HS type 2 in eleven patients. Clinical features, postoperative seizure outcomes, hippocampal subfields volumes, fractional anisotropy (FA) values of white matter regions and graph theory parameters were explored and compared between the HS type 1 and HS type 2 groups. RESULTS: No significant differences in clinical features and postsurgical seizure outcomes were found between the HS type 1 and type 2 groups. However, the HS type 1 group showed extra atrophy in ipsilateral parasubiculum than healthy controls and more severe atrophy in contralateral hippocampal fissure than the HS type 2 group. More extensive FA decrease were also observed in the HS type 1 group, involving ipsilateral optic radiation, superior fronto-occipital fasciculus, contralateral uncinate fasciculus, tapetum, bilateral hippocampal cingulum, corona radiata, etc. Furthermore, in spite of similar impairments in characteristic path length, global efficiency and local efficiency in two HS groups, the HS type 1 group showed additional decrease of clustering coefficient than healthy controls. CONCLUSIONS: HS type 1 and 2 groups had similar clinical characteristics and postoperative seizure outcomes. More widespread neuronal cell loss in the HS type 1 group contributed to more extensive structural damage and connectivity abnormality. These results shed new light on the imaging correlates of different HS pathology.

Different brain functional network abnormalities between attention-deficit/hyperactivity disorder youth with and without familial risk for bipolar disorder.

Attention-deficit/hyperactivity disorder (ADHD) commonly precedes the initial onset of mania in youth with familial risk for bipolar disorder (BD). Although ADHD youth with and without BD familial risk exhibit different clinical features, associated neuropathophysiological mechanisms remain poorly understood. This study aimed to identify brain functional network abnormalities associated with ADHD in youth with and without familial risk for BD. Resting-state functional magnetic resonance imaging scans were acquired from 37 ADHD youth with a family history of BD (high-risk), 45 ADHD youth without a family history of BD (low-risk), and 32 healthy controls (HC). Individual whole-brain functional networks were constructed, and graph theory analysis was applied to estimate network topological metrics. Topological metrics, including network efficiency, small-worldness and nodal centrality, were compared across groups, and associations between topological metrics and clinical ratings were evaluated. Compared to HC, low-risk ADHD youth exhibited weaker global integration (i.e., decreased global efficiency and increased characteristic path length), while high-risk ADHD youth showed a disruption of localized network components with decreased frontoparietal and frontolimbic connectivity. Common topological deficits were observed in the medial superior frontal gyrus between low- and high-risk ADHD. Distinct network deficits were found in the inferior parietal lobule and corticostriatal circuitry. Associations between global topological metrics and externalizing symptoms differed significantly between the two ADHD groups. Different patterns of functional network topological abnormalities were found in high- as compared to low-risk ADHD, suggesting that ADHD in youth with BD familial risk may represent a phenotype that is different from ADHD alone.

Morphological abnormalities in youth with bipolar disorder and their relationship to clinical characteristics.

OBJECTIVES: To characterize the neuroanatomy of BD in youth and its correlation to clinical characteristics. METHODS: The current study includes a sample of 105 unmedicated youth with first-episode BD, aged between 10.1 and 17.9 years, and 61 healthy comparison adolescents, aged between 10.1 and 17.7 years, who were matched for age, race, sex, socioeconomic status, intelligence quotient (IQ), and education level. T1-weighted magnetic resonance imaging (MRI) images were obtained using a 4 T MRI scanner. Freesurfer (V6.0) was used to preprocess and parcellate the structural data, and 68 cortical and 12 subcortical regions were considered for statistical comparisons. The relationship between morphological deficits and clinical and demographic characteristics were evaluated using linear models. RESULTS: Compared with healthy youth, youth with BD had decreased cortical thickness in frontal, parietal, and anterior cingulate regions. These youth also showed decreased gray matter volumes in 6 of the 12 subcortical regions examined including thalamus, putamen, amygdala and caudate. In further subgroup analyses, we found that youth with BD with comorbid attention-deficit hyperactivity disorder (ADHD) or with psychotic symptoms had more significant deficits in subcortical gray matter volume. LIMITATIONS: We cannot provide information about the course of structural changes and impact of treatment and illness progression. CONCLUSIONS: Our findings indicate that youth with BD have significant neurostructural deficits in both cortical and subcortical regions mainly located in the regions related to emotion processing and regulation. Variability in clinical characteristics and comorbidities may contribute to the severity of anatomic alterations in this disorder.

Regional microstructural differences in ADHD youth with and without a family history of bipolar I disorder.

BACKGROUND: Having a first-degree relative with bipolar I disorder (BD) in conjunction with prodromal attention deficit/hyperactivity disorder (ADHD) may represent a unique phenotype that confers greater risk for developing BD than ADHD alone. However, underlying neuropathoetiological mechanisms remain poorly understood. This cross-sectional study compared regional microstructure in psychostimulant-free ADHD youth with ('high-risk', HR) and without ('low-risk', LR) a first-degree relative with BD, and healthy controls (HC). METHODS: A total of 140 (high-risk, n = 44; low-risk, n = 49; and HC, n = 47) youth (mean age: 14.1 ± 2.5 years, 65 % male) were included in the analysis. Diffusion tensor images were collected and fractional anisotropy (FA) and mean diffusivity (MD) maps were calculated. Both tract-based and voxel-based analyses were performed. Correlations between clinical ratings and microstructural metrics that differed among groups were examined. RESULTS: No significant group differences in major long-distance fiber tracts were observed. The high-risk ADHD group exhibited predominantly higher FA and lower MD in frontal, limbic, and striatal subregions compared with the low-risk ADHD group. Both low-risk and high-risk ADHD groups exhibited higher FA in unique and overlapping regions compared with HC subjects. Significant correlations between regional microstructural metrics and clinical ratings were observed in ADHD groups. LIMITATIONS: Prospective longitudinal studies will be required to determine the relevance of these findings to BD risk progression. CONCLUSIONS: Psychostimulant-free ADHD youth with a BD family history exhibit different microstructure alterations in frontal, limbic, and striatal regions compared with ADHD youth without a BD family history, and may therefore represent a unique phenotype relevant to BD risk progression.

Predicting acupuncture efficacy for functional dyspepsia based on functional brain network features: a machine learning study.

Acupuncture is effective in treating functional dyspepsia (FD), while its efficacy varies significantly from different patients. Predicting the responsiveness of different patients to acupuncture treatment based on the objective biomarkers would assist physicians to identify the candidates for acupuncture therapy. One hundred FD patients were enrolled, and their clinical characteristics and functional brain MRI data were collected before and after treatment. Taking the pre-treatment functional brain network as features, we constructed the support vector machine models to predict the responsiveness of FD patients to acupuncture treatment. These features contributing critically to the accurate prediction were identified, and the longitudinal analyses of these features were performed on acupuncture responders and non-responders. Results demonstrated that prediction models achieved an accuracy of 0.76 ± 0.03 in predicting acupuncture responders and non-responders, and a R2 of 0.24 ± 0.02 in predicting dyspeptic symptoms relief. Thirty-eight functional brain network features associated with the orbitofrontal cortex, caudate, hippocampus, and anterior insula were identified as the critical predictive features. Changes in these predictive features were more pronounced in responders than in non-responders. In conclusion, this study provided a promising approach to predicting acupuncture efficacy for FD patients and is expected to facilitate the optimization of personalized acupuncture treatment plans for FD.

Brain morphometric features predict medication response in youth with bipolar disorder: a prospective randomized clinical trial.

BACKGROUND: Identification of treatment-specific predictors of drug therapies for bipolar disorder (BD) is important because only about half of individuals respond to any specific medication. However, medication response in pediatric BD is variable and not well predicted by clinical characteristics. METHODS: A total of 121 youth with early course BD (acute manic/mixed episode) were prospectively recruited and randomized to 6 weeks of double-blind treatment with quetiapine (n = 71) or lithium (n = 50). Participants completed structural magnetic resonance imaging (MRI) at baseline before treatment and 1 week after treatment initiation, and brain morphometric features were extracted for each individual based on MRI scans. Positive antimanic treatment response at week 6 was defined as an over 50% reduction of Young Mania Rating Scale scores from baseline. Two-stage deep learning prediction model was established to distinguish responders and non-responders based on different feature sets. RESULTS: Pre-treatment morphometry and morphometric changes occurring during the first week can both independently predict treatment outcome of quetiapine and lithium with balanced accuracy over 75% (all p < 0.05). Combining brain morphometry at baseline and week 1 allows prediction with the highest balanced accuracy (quetiapine: 83.2% and lithium: 83.5%). Predictions in the quetiapine and lithium group were found to be driven by different morphometric patterns. CONCLUSIONS: These findings demonstrate that pre-treatment morphometric measures and acute brain morphometric changes can serve as medication response predictors in pediatric BD. Brain morphometric features may provide promising biomarkers for developing biologically-informed treatment outcome prediction and patient stratification tools for BD treatment development.

Effects of short-term quetiapine and lithium therapy for acute manic or mixed episodes on the limbic system and emotion regulation circuitry in youth with bipolar disorder.

Disruptions in the limbic system, and in emotion regulation circuitry that supports affect modulation, have been reported during acute manic episodes of bipolar disorder (BD). The impact of pharmacological treatment on these deficits, especially in youth, remains poorly characterized. 107 youths with acute manic or mixed episodes of bipolar I disorder and 60 group-matched healthy controls were recruited. Youth with bipolar disorder were randomized to double-blind treatment with quetiapine or lithium and assessed weekly. Task-based fMRI studies were performed using an identical pairs continuous performance task (CPT-IP) at pre-treatment baseline and post-treatment weeks one and six. Region of interest analyses focused on the limbic system and ventral PFC - basal ganglia - thalamocortical loop structures known to be involved in emotion regulation. Changes in regional activation were compared between the two treatment groups, and pretreatment regional activation was used to predict treatment outcome. Mania treatment scores improved more rapidly in the quetiapine than lithium treated group, as did significant normalization of neural activation toward that of healthy individuals in left amygdala (p = 0.007), right putamen (p < 0.001), and right globus pallidus (p = 0.003). Activation changes in the right putamen were correlated with reduction of mania symptoms. The limbic and emotion regulation system activation at baseline and week one predicted treatment outcome in youth with bipolar disorder with significant accuracy (up to 87.5%). Our findings document more rapid functional brain changes associated with quetiapine than lithium treatment in youth with bipolar disorder, with most notable changes in the limbic system and emotion regulation circuitry. Pretreatment alterations in these regions predicted treatment response. These findings advance understanding of regional brain alterations in youth with bipolar disorder, and show that fMRI data can predict treatment outcome before it can be determined clinically, highlighting the potential utility of fMRI biomarkers for early prediction of treatment outcomes in bipolar disorder.Clinical Trials Registration: Name: Multimodal Neuroimaging of Treatment Effects in Adolescent Mania. URL: https://clinicaltrials.gov/ . Registration number: NCT00893581.

Altered Resting State Networks Before and After Temporal Lobe Epilepsy Surgery.

OBJECTIVES: To explore the resting state networks (RSNs) alterations in patients with unilateral mesial temporal lobe epilepsy (mTLE) before and after successful surgery. METHODS: Resting-state functional MRI and T1-weighted structural MRI were obtained in 37 mTLE patients who achieved seizure freedom after anterior temporal lobectomy. Patients were scanned before surgery and at two years after surgery. Twenty-eight age- and sex-matched healthy controls were scanned once. Functional connectivity (FC) changes within and between ten common RSNs before and after surgery, and FC changes between hippocampus and RSNs were explored. RESULTS: Before surgery, decreased FC was found within visual network and basal ganglia network, while after surgery, FC within basal ganglia network further decreased but FC within sensorimotor network and dorsal attention network increased. Before surgery, between-network FC related to basal ganglia network, visual network and dorsal attention network decreased, while between-network FC related to default mode network increased. After surgery, between-network FC related to visual network and dorsal attention network significantly increased. In addition, before surgery, ipsilateral hippocampus showed decreased FC with visual network, basal ganglia network, sensorimotor network, default mode network and frontoparietal network, while contralateral rostral hippocampus showed increased FC with salience network. After surgery, no obvious FC changes were found between contralateral hippocampus and these RSNs. CONCLUSION: MTLE patients showed significant RSNs alterations before and after surgery. Basal ganglia network showed progressive decline in functional connectivity. Successful surgery may lead to RSNs reorganization. These results provide preliminary evidence for postoperative functional remodeling at whole-brain-network level.

Altered dynamic functional connectivity of striatal-cortical circuits in Juvenile Myoclonic Epilepsy.

OBJECTIVE: To investigate whether the dynamic functional connectivity (dFC) of striatal-cortical circuits changes in juvenile myoclonic epilepsy (JME). METHODS: The resting-state EEG-fMRI and the sliding-window approach were adopted to explore the dynamic striatal-cortical circuitry in thirty JME patients compared with 30 well-matched health controls (HCs). Six pairs of striatal seeds were selected as regions of interests. The correlation analysis was performed to reveal the relationship between the altered dFC variability and clinical variables in JME group. RESULTS: JME patients exhibited increased dFC variability mainly involved in fronto-striatal and striatal-thalamic networks; decreased dFC variability between striatum subdivisions and default mode network (DMN) regions compared with HCs (p<0.05, GRF corrected). In addition, the hypervariability between left ventral-rostral putamen and left medial superior frontal gyrus was positively (r= 0.493, p=0.008) correlated with the mean frequency score of myoclonic seizures in JME group. CONCLUSION: JME presented altered dFC variability in striatal-cortical circuits. The pattern of altered circuits showed increased variability in fronto-striatal and striatal-thalamic networks and decreased variability in striatal-DMN. These results provide novel information about the dynamic neural striatal-cortical circuitry of JME.

GMILT: A Novel Transformer Network That Can Noninvasively Predict EGFR Mutation Status.

Noninvasively and accurately predicting the epidermal growth factor receptor (EGFR) mutation status is a clinically vital problem. Moreover, further identifying the most suspicious area related to the EGFR mutation status can guide the biopsy to avoid false negatives. Deep learning methods based on computed tomography (CT) images may improve the noninvasive prediction of EGFR mutation status and potentially help clinicians guide biopsies by visual methods. Inspired by the potential inherent links between EGFR mutation status and invasiveness information, we hypothesized that the predictive performance of a deep learning network can be improved through extra utilization of the invasiveness information. Here, we created a novel explainable transformer network for EGFR classification named gated multiple instance learning transformer (GMILT) by integrating multi-instance learning and discriminative weakly supervised feature learning. Pathological invasiveness information was first introduced into the multitask model as embeddings. GMILT was trained and validated on a total of 512 patients with adenocarcinoma and tested on three datasets (the internal test dataset, the external test dataset, and The Cancer Imaging Archive (TCIA) public dataset). The performance (area under the curve (AUC) = 0.772 on the internal test dataset) of GMILT exceeded that of previously published methods and radiomics-based methods (i.e., random forest and support vector machine) and attained a preferable generalization ability (AUC = 0.856 in the TCIA test dataset and AUC = 0.756 in the external dataset). A diameter-based subgroup analysis further verified the efficiency of our model (most of the AUCs exceeded 0.772) to noninvasively predict EGFR mutation status from computed tomography (CT) images. In addition, because our method also identified the "core area" of the most suspicious area related to the EGFR mutation status, it has the potential ability to guide biopsies.

Topologically Disrupted Gray Matter Networks in Drug-Naïve Essential Tremor Patients With Poor Sleep Quality.

Background: Sleep disturbances are widespread among patients with essential tremor (ET) and may have adverse effects on patients' quality of life. However, the pathophysiology underlying poor quality of sleep (QoS) in patients with ET remains unclear. Our study aimed to identify gray matter (GM) network alterations in the topological properties of structural MRI related to QoS in patients with ET. Method: We enrolled 45 ET patients with poor QoS (SleET), 59 ET patients with normal QoS (NorET), and 66 healthy controls (HC), and they all underwent a three-dimensional T1-weighted MRI scan. We used a graph-theoretical approach to investigate the topological organization of GM morphological networks, and individual morphological brain networks were constructed according to the interregional similarity of GM volume distributions. Furthermore, we performed network-based statistics, and partial correlation analyses between topographic features and clinical characteristics were conducted. Results: Global network organization was disrupted in patients with ET. Compared with the NorET group, the SleET group exhibited disrupted topological GM network organization with a shift toward randomization. Moreover, they showed altered nodal centralities in mainly the frontal, temporal, parietal, and cerebellar lobes. Morphological connection alterations within the default mode network (DMN), salience, and basal ganglia networks were observed in the SleET group and were generally more extensive than those in the NorET and HC groups. Alterations within the cerebello-thalamo-(cortical) network were only detected in the SleET group. The nodal degree of the left thalamus was negatively correlated with the Fahn-Tolosa-Marin Tremor Rating Scale score (r = -0.354, p =0.027). Conclusion: Our findings suggest that potential complex interactions underlie tremor and sleep disruptions in patients with ET. Disruptions within the DMN and the cerebello-thalamo-(cortical) network may have a broader impact on sleep quality in patients with ET. Our results offer valuable insight into the neural mechanisms underlying poor QoS in patients with ET.

Graph Convolutional Networks Reveal Network-Level Functional Dysconnectivity in Schizophrenia.

BACKGROUND AND HYPOTHESIS: Schizophrenia is increasingly understood as a disorder of brain dysconnectivity. Recently, graph-based approaches such as graph convolutional network (GCN) have been leveraged to explore complex pairwise similarities in imaging features among brain regions, which can reveal abstract and complex relationships within brain networks. STUDY DESIGN: We used GCN to investigate topological abnormalities of functional brain networks in schizophrenia. Resting-state functional magnetic resonance imaging data were acquired from 505 individuals with schizophrenia and 907 controls across 6 sites. Whole-brain functional connectivity matrix was extracted for each individual. We examined the performance of GCN relative to support vector machine (SVM), extracted the most salient regions contributing to both classification models, investigated the topological profiles of identified salient regions, and explored correlation between nodal topological properties of each salient region and severity of symptom. STUDY RESULTS: GCN enabled nominally higher classification accuracy (85.8%) compared with SVM (80.9%). Based on the saliency map, the most discriminative brain regions were located in a distributed network including striatal areas (ie, putamen, pallidum, and caudate) and the amygdala. Significant differences in the nodal efficiency of bilateral putamen and pallidum between patients and controls and its correlations with negative symptoms were detected in post hoc analysis. CONCLUSIONS: The present study demonstrates that GCN allows classification of schizophrenia at the individual level with high accuracy, indicating a promising direction for detection of individual patients with schizophrenia. Functional topological deficits of striatal areas may represent a focal neural deficit of negative symptomatology in schizophrenia.