Brain regions with abnormal network properties in severe epilepsy of Lennox-Gastaut phenotype: Multivariate analysis of task-free fMRI

Abnormal global and local connectivity in patients with anti-N-methyl-D-aspartate receptor encephalitis: A resting-state functional MRI study

OBJECTIVE

We decided to investigate the changes of global and local connectivity in anti-N-methyl-D-aspartate receptor (anti-NMDAR) encephalitis patients based on eigenvector centrality (EC) and regional homogeneity (ReHo). We sought new biomarkers to identify the patients based on multivariate pattern analysis (MVPA).

METHODS

Functional MRI (fMRI) was performed on all participants. EC, ReHo and MVPA were used to analyze the fMRI images. The correlation between the global or local connectivity and neuropsychology tests was detected.

RESULTS

The MoCA scores of the patients were lower than those of the healthy controls (HCs), while the HAMD24 and HAMA scores of the patients were higher than those of the HCs. Increased EC values in the right calcarine (CAL.R) and decreased EC values in the right putamen (PUT.R) distinguished these subjects with anti-NMDAR encephalitis from HCs. The higher ReHo values in the left postcentral gyrus (PoCG.L) were detected in the patients. The correlation analysis showed that the EC values in the PUT.R were negatively correlated with HAMD24 and HAMA scores, while the ReHo values in the PoCG.L were negatively correlated with MoCA scores. Better classification performance was reached in the EC-based classifier (AUC = 0.80), while weaker classification performance was achieved in the ReHo-based classifier (AUC = 0.74) or the classifier based on EC and ReHo (AUC = 0.74). The brain areas with large weights were located in the frontal lobe, parietal lobe, cerebellum and basal ganglia.

CONCLUSION

Our findings suggest that abnormal global and local connectivity may play an important part in the pathophysiological mechanism of neuropsychiatric symptoms in the anti-NMDAR encephalitis patients. The EC-based classifier may be better than the ReHo-based classifier in identifying anti-NMDAR encephalitis patients.

Artificial intelligence in epilepsy - applications and pathways to the clinic

Artificial intelligence (AI) is rapidly transforming health care, and its applications in epilepsy have increased exponentially over the past decade. Integration of AI into epilepsy management promises to revolutionize the diagnosis and treatment of this complex disorder. However, translation of AI into neurology clinical practice has not yet been successful, emphasizing the need to consider progress to date and assess challenges and limitations of AI. In this Review, we provide an overview of AI applications that have been developed in epilepsy using a variety of data modalities: neuroimaging, electroencephalography, electronic health records, medical devices and multimodal data integration. For each, we consider potential applications, including seizure detection and prediction, seizure lateralization, localization of the seizure-onset zone and assessment for surgical or neurostimulation interventions, and review the performance of AI tools developed to date. We also discuss methodological considerations and challenges that must be addressed to successfully integrate AI into clinical practice. Our goal is to provide an overview of the current state of the field and provide guidance for leveraging AI in future to improve management of epilepsy.

Sampling inequalities affect generalization of neuroimaging-based diagnostic classifiers in psychiatry

BACKGROUND

The development of machine learning models for aiding in the diagnosis of mental disorder is recognized as a significant breakthrough in the field of psychiatry. However, clinical practice of such models remains a challenge, with poor generalizability being a major limitation.

METHODS

Here, we conducted a pre-registered meta-research assessment on neuroimaging-based models in the psychiatric literature, quantitatively examining global and regional sampling issues over recent decades, from a view that has been relatively underexplored. A total of 476 studies (n = 118,137) were included in the current assessment. Based on these findings, we built a comprehensive 5-star rating system to quantitatively evaluate the quality of existing machine learning models for psychiatric diagnoses.

RESULTS

A global sampling inequality in these models was revealed quantitatively (sampling Gini coefficient (G) = 0.81, p < .01), varying across different countries (regions) (e.g., China, G = 0.47; the USA, G = 0.58; Germany, G = 0.78; the UK, G = 0.87). Furthermore, the severity of this sampling inequality was significantly predicted by national economic levels (β =  - 2.75, p < .001, R²_adj = 0.40; r =  - .84, 95% CI: - .41 to - .97), and was plausibly predictable for model performance, with higher sampling inequality for reporting higher classification accuracy. Further analyses showed that lack of independent testing (84.24% of models, 95% CI: 81.0-87.5%), improper cross-validation (51.68% of models, 95% CI: 47.2-56.2%), and poor technical transparency (87.8% of models, 95% CI: 84.9-90.8%)/availability (80.88% of models, 95% CI: 77.3-84.4%) are prevailing in current diagnostic classifiers despite improvements over time. Relating to these observations, model performances were found decreased in studies with independent cross-country sampling validations (all p < .001, BF₁₀ > 15). In light of this, we proposed a purpose-built quantitative assessment checklist, which demonstrated that the overall ratings of these models increased by publication year but were negatively associated with model performance.

CONCLUSIONS

Together, improving sampling economic equality and hence the quality of machine learning models may be a crucial facet to plausibly translating neuroimaging-based diagnostic classifiers into clinical practice.

Evaluation of Risk of Bias in Neuroimaging-Based Artificial Intelligence Models for Psychiatric Diagnosis: A Systematic Review

IMPORTANCE

Neuroimaging-based artificial intelligence (AI) diagnostic models have proliferated in psychiatry. However, their clinical applicability and reporting quality (ie, feasibility) for clinical practice have not been systematically evaluated.

OBJECTIVE

To systematically assess the risk of bias (ROB) and reporting quality of neuroimaging-based AI models for psychiatric diagnosis.

EVIDENCE REVIEW

PubMed was searched for peer-reviewed, full-length articles published between January 1, 1990, and March 16, 2022. Studies aimed at developing or validating neuroimaging-based AI models for clinical diagnosis of psychiatric disorders were included. Reference lists were further searched for suitable original studies. Data extraction followed the CHARMS (Checklist for Critical Appraisal and Data Extraction for Systematic Reviews of Prediction Modeling Studies) and PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) guidelines. A closed-loop cross-sequential design was used for quality control. The PROBAST (Prediction Model Risk of Bias Assessment Tool) and modified CLEAR (Checklist for Evaluation of Image-Based Artificial Intelligence Reports) benchmarks were used to systematically evaluate ROB and reporting quality.

FINDINGS

A total of 517 studies presenting 555 AI models were included and evaluated. Of these models, 461 (83.1%; 95% CI, 80.0%-86.2%) were rated as having a high overall ROB based on the PROBAST. The ROB was particular high in the analysis domain, including inadequate sample size (398 of 555 models [71.7%; 95% CI, 68.0%-75.6%]), poor model performance examination (with 100% of models lacking calibration examination), and lack of handling data complexity (550 of 555 models [99.1%; 95% CI, 98.3%-99.9%]). None of the AI models was perceived to be applicable to clinical practices. Overall reporting completeness (ie, number of reported items/number of total items) for the AI models was 61.2% (95% CI, 60.6%-61.8%), and the completeness was poorest for the technical assessment domain with 39.9% (95% CI, 38.8%-41.1%).

CONCLUSIONS AND RELEVANCE

This systematic review found that the clinical applicability and feasibility of neuroimaging-based AI models for psychiatric diagnosis were challenged by a high ROB and poor reporting quality. Particularly in the analysis domain, ROB in AI diagnostic models should be addressed before clinical application.

Norwegian population-based study of long-term effects, safety, and predictors of response of vagus nerve stimulation treatment in drug-resistant epilepsy: The NORPulse study

OBJECTIVE

This study was undertaken to evaluate the efficacy of vagus nerve stimulation (VNS) over time, and to determine which patient groups derive the most benefit.

METHODS

Long-term outcomes are reported in 436 epilepsy patients from a VNS quality registry (52.8% adults, 47.2% children), with a median follow-up of 75 months. Patients were stratified according to evolution of response into constant responders, fluctuating responders, and nonresponders. The effect was evaluated at 6, 12, 24, 36, and 60 months. Multivariate regression analysis was used to identify predictors of response.

RESULTS

The cumulative probability of ≥50% seizure reduction was 60%; however, 15% of patients showed a fluctuating course. Of those becoming responders, 89.5% (230/257) did so within 2 years. A steady increase in effect was observed among constant responders, with 48.7% (19/39) of those becoming seizure-free and 29.3% (39/133) with ≥75% seizure reduction achieving these effects within 2-5 years. Some effect (25%-<50%) at 6 months was a positive predictor of becoming a responder (odds ratio [OR] = 10.18, p < .0001) and having ≥75% reduction at 2 years (OR = 3.34, p = .03). Patients without intellectual disability had ORs of 3.34 and 3.11 of having ≥75% reduction at 2 and 5 years, respectively, and an OR of 6.22 of being seizure-free at last observation. Patients with unchanged antiseizure medication over the observation period showed better responder rates at 2 (63.0% vs. 43.1%, p = .002) and 5 years (63.4% vs. 46.3%, p = .031) than patients whose antiseizure medication was modified. Responder rates were higher for posttraumatic (70.6%, p = .048) and poststroke epilepsies (75.0%, p = .05) than other etiologies (46.5%).

SIGNIFICANCE

Our data indicate that the effect of VNS increases over time and that there are important clinical decision points at 6 and 24 months for evaluating and adjusting the treatment. There should be better selection of candidates, as certain patient groups and epilepsy etiologies respond more favorably.

Clinical Application of Machine Learning Models for Brain Imaging in Epilepsy: A Review

Epilepsy is a common neurological disorder characterized by recurrent and disabling seizures. An increasing number of clinical and experimental applications of machine learning (ML) methods for epilepsy and other neurological and psychiatric disorders are available. ML methods have the potential to provide a reliable and optimal performance for clinical diagnoses, prediction, and personalized medicine by using mathematical algorithms and computational approaches. There are now several applications of ML for epilepsy, including neuroimaging analyses. For precise and reliable clinical applications in epilepsy and neuroimaging, the diverse ML methodologies should be examined and validated. We review the clinical applications of ML models for brain imaging in epilepsy obtained from a PubMed database search in February 2021. We first present an overview of typical neuroimaging modalities and ML models used in the epilepsy studies and then focus on the existing applications of ML models for brain imaging in epilepsy based on the following clinical aspects: (i) distinguishing individuals with epilepsy from healthy controls, (ii) lateralization of the temporal lobe epilepsy focus, (iii) the identification of epileptogenic foci, (iv) the prediction of clinical outcomes, and (v) brain-age prediction. We address the practical problems and challenges described in the literature and suggest some future research directions.

Characteristics of Genetic Variations Associated With Lennox-Gastaut Syndrome in Korean Families

Lennox-Gastaut syndrome (LGS) is a severe type of childhood-onset epilepsy characterized by multiple types of seizures, specific discharges on electroencephalography, and intellectual disability. Most patients with LGS do not respond well to drug treatment and show poor long-term prognosis. Approximately 30% of patients without brain abnormalities have unidentifiable causes. Therefore, accurate diagnosis and treatment of LGS remain challenging. To identify causative mutations of LGS, we analyzed the whole-exome sequencing data of 17 unrelated Korean families, including patients with LGS and LGS-like epilepsy without brain abnormalities, using the Genome Analysis Toolkit. We identified 14 mutations in 14 genes as causes of LGS or LGS-like epilepsy. 64 percent of the identified genes were reported as LGS or epilepsy-related genes. Many of these variations were novel and considered as pathogenic or likely pathogenic. Network analysis was performed to classify the identified genes into two network clusters: neuronal signal transmission or neuronal development. Additionally, knockdown of two candidate genes with insufficient evidence of neuronal functions, SLC25A39 and TBC1D8, decreased neurite outgrowth and the expression level of MAP2, a neuronal marker. These results expand the spectrum of genetic variations and may aid the diagnosis and management of individuals with LGS.

Machine learning in mental health: a scoping review of methods and applications

BACKGROUND

This paper aims to synthesise the literature on machine learning (ML) and big data applications for mental health, highlighting current research and applications in practice.

METHODS

We employed a scoping review methodology to rapidly map the field of ML in mental health. Eight health and information technology research databases were searched for papers covering this domain. Articles were assessed by two reviewers, and data were extracted on the article's mental health application, ML technique, data type, and study results. Articles were then synthesised via narrative review.

RESULTS

Three hundred papers focusing on the application of ML to mental health were identified. Four main application domains emerged in the literature, including: (i) detection and diagnosis; (ii) prognosis, treatment and support; (iii) public health, and; (iv) research and clinical administration. The most common mental health conditions addressed included depression, schizophrenia, and Alzheimer's disease. ML techniques used included support vector machines, decision trees, neural networks, latent Dirichlet allocation, and clustering.

CONCLUSIONS

Overall, the application of ML to mental health has demonstrated a range of benefits across the areas of diagnosis, treatment and support, research, and clinical administration. With the majority of studies identified focusing on the detection and diagnosis of mental health conditions, it is evident that there is significant room for the application of ML to other areas of psychology and mental health. The challenges of using ML techniques are discussed, as well as opportunities to improve and advance the field.

The dynamics of functional connectivity in neocortical focal epilepsy

Focal epilepsy is characterised by paroxysmal events, reflecting changes in underlying local brain networks. To capture brain network activity at the maximal temporal resolution of the acquired functional magnetic resonance imaging (fMRI) data, we have previously developed a novel analysis framework called Dynamic Regional Phase Synchrony (DRePS). DRePS measures instantaneous mean phase coherence within neighbourhoods of brain voxels. We use it here to examine how the dynamics of the functional connections of regional brain networks are altered in neocortical focal epilepsy. Using task-free fMRI data from 21 subjects with focal epilepsy and 21 healthy controls, we calculated the power spectral density of DRePS, which is a measure of signal variability in local connectivity estimates. Whole-brain averaged power spectral density of DRePS, or signal variability of local connectivity, was significantly higher in epilepsy subjects compared to healthy controls. Maximal increase in DRePS spectral power was seen in bilateral inferior frontal cortices, ipsilateral mid-cingulate gyrus, superior temporal gyrus, caudate head, and contralateral cerebellum. Our results provide further evidence of common brain abnormalities across people with focal epilepsy. We postulate that dynamic changes in specific cortical brain areas may help maintain brain function in the presence of pathological epileptiform network activity in neocortical focal epilepsy.

•

Focal epilepsy is a disease defined by its inherent paroxysmal brain changes.

•

However, most fMRI connectivity methods do not capture dynamic brain network activity.

•

We use our recently developed method, DRePS, to identify abnormal dynamic brain networks in focal epilepsy.

•

Fluctuations of dynamic brain network activity were increased in the frontal-temporal cortex, cingulum and cerebellum, in focal epilepsy.

•

These regions may be involved in (re-)gaining cortical homeostasis in an otherwise abnormal brain.

Altered degree centrality in childhood absence epilepsy: A resting-state fMRI study

Modern network studies have suggested that the pathology of many neurological diseases is in fact not equally distributed over the brain but preferentially affects the hub regions. This study aims to investigate how hub regions were affected in Children with Childhood absence epilepsy (CAE) using resting-state fMRI (rs-fMRI). As one important measures obtained from rs-fMRI, degree centrality (DC) calculates the number of direct connections between a given node and the rest of the brain within the entire connectivity matrix of the brain. In this study, twenty-five CAE children and 25 healthy controls were recruited to investigate the DC changes in CAE patients. Compared with healthy controls, children with CAE showed significantly decreased DC in default mode network (DMN, medial prefrontal cortex, posterior cingulate cortex, precuneus and middle temporal cortex) and increased DC in thalamus. Importantly, significant negative correlation between the epilepsy duration and DC was found in precuneus. Our results suggested selective and specific disruption of hub nodes, especially thalamus and the highly connected brain regions within DMN, might underlie the pathophysiological mechanism of CAE.

A Hitchhiker's Guide to Functional Magnetic Resonance Imaging

Functional Magnetic Resonance Imaging (fMRI) studies have become increasingly popular both with clinicians and researchers as they are capable of providing unique insights into brain functions. However, multiple technical considerations (ranging from specifics of paradigm design to imaging artifacts, complex protocol definition, and multitude of processing and methods of analysis, as well as intrinsic methodological limitations) must be considered and addressed in order to optimize fMRI analysis and to arrive at the most accurate and grounded interpretation of the data. In practice, the researcher/clinician must choose, from many available options, the most suitable software tool for each stage of the fMRI analysis pipeline. Herein we provide a straightforward guide designed to address, for each of the major stages, the techniques, and tools involved in the process. We have developed this guide both to help those new to the technique to overcome the most critical difficulties in its use, as well as to serve as a resource for the neuroimaging community.