Evaluation of machine learning algorithms for treatment outcome prediction in patients with epilepsy based on structural connectome data

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.

Structural connectivity as a predictive factor for perampanel response in patients with epilepsy

OBJECTIVES

This study aimed to identify clinical characteristics that could predict the response to perampanel (PER) and determine whether structural connectivity is a predictive factor.

METHODS

We enrolled patients with epilepsy who received PER and were followed-up for a minimum of 12 months. Good PER responders, who were seizure-free or presented with more than 50 % seizure reduction, were classified separately from poor PER responders who had seizure reduction of less than 50 % or non-responders. A graph theoretical analysis was conducted based on diffusion tensor imaging to calculate network measures of structural connectivity among patients with epilepsy.

RESULTS

106 patients with epilepsy were enrolled, including 26 good PER responders and 80 poor PER responders. Good PER responders used fewer anti-seizure medications before PER administration compared to those by poor PER responders (3 vs. 4; p = 0.042). Early PER treatment was more common in good PER responders than poor PER responders (46.2 vs. 21.3 %, p = 0.014). Regarding cortical structural connectivity, the global efficiency was higher and characteristic path length was lower in good PER responders than in poor PER responders (0.647 vs. 0.635, p = 0.006; 1.726 vs. 1,759, p = 0.008, respectively). For subcortical structural connectivity, the mean clustering coefficient and small-worldness index were higher in good PER responders than in poor PER responders (0.821 vs. 0.791, p = 0.009; 0.597 vs. 0.560, p = 0.009, respectively).

CONCLUSION

This study demonstrated that early PER administration can predict a good PER response in patients with epilepsy, and structural connectivity could potentially offer clinical utility in predicting PER response.

The use of automated and AI-driven algorithms for the detection of hippocampal sclerosis and focal cortical dysplasia

In drug-resistant epilepsy, magnetic resonance imaging (MRI) plays a central role in detecting lesions as it offers unmatched spatial resolution and whole-brain coverage. In addition, the last decade has witnessed continued developments in MRI-based computer-aided machine-learning techniques for improved diagnosis and prognosis. In this review, we focus on automated algorithms for the detection of hippocampal sclerosis and focal cortical dysplasia, particularly in cases deemed as MRI negative, with an emphasis on studies with histologically validated data. In addition, we discuss imaging-derived prognostic markers, including response to anti-seizure medication, post-surgical seizure outcome, and cognitive reserves. We also highlight the advantages and limitations of these approaches and discuss future directions toward person-centered care.

Predictive models of epilepsy outcomes

PURPOSE OF REVIEW

Multiple complex medical decisions are necessary in the course of a chronic disease like epilepsy. Predictive tools to assist physicians and patients in navigating this complexity have emerged as a necessity and are summarized in this review.

RECENT FINDINGS

Nomograms and online risk calculators are user-friendly and offer individualized predictions for outcomes ranging from safety of antiseizure medication withdrawal (accuracy 65-73%) to seizure-freedom, naming, mood, and language outcomes of resective epilepsy surgery (accuracy 72-81%). Improving their predictive performance is limited by the nomograms' inability to ingest complex data inputs. Conversely, machine learning offers the potential of multimodal and expansive model inputs achieving human-expert level accuracy in automated scalp electroencephalogram (EEG) interpretation but lagging in predictive performance or requiring validation for other applications.

SUMMARY

Good to excellent predictive models are now available to guide medical and surgical epilepsy decision-making with nomograms offering individualized predictions and user-friendly tools, and machine learning approaches offering the potential of improved performance. Future research is necessary to bridge the two approaches for optimal translation to clinical care.

Epileptogenic networks in extra temporal lobe epilepsy

Extra temporal lobe epilepsy (eTLE) may involve heterogenous widespread cerebral networks. We investigated the structural network of an eTLE cohort, at the postulated epileptogenic zone later surgically removed, as a network node: the resection zone (RZ). We hypothesized patients with an abnormal connection to/from the RZ to have proportionally increased abnormalities based on topological proximity to the RZ, in addition to poorer post-operative seizure outcome. Structural and diffusion MRI were collected for 22 eTLE patients pre- and post-surgery, and for 29 healthy controls. The structural connectivity of the RZ prior to surgery, measured via generalized fractional anisotropy (gFA), was compared with healthy controls. Abnormal connections were identified as those with substantially reduced gFA (z < -1.96). For patients with one or more abnormal connections to/from the RZ, connections with closer topological distance to the RZ had higher proportion of abnormalities. The minority of the seizure-free patients (3/11) had one or more abnormal connections, while most non-seizure-free patients (8/11) had abnormal connections to the RZ. Our data suggest that eTLE patients with one or more abnormal structural connections to/from the RZ had more proportional abnormal connections based on topological distance to the RZ and associated with reduced chance of seizure freedom post-surgery.

Learning algorithms estimate pose and detect motor anomalies in flies exposed to minimal doses of a toxicant

Pesticide exposure, even at low doses, can have detrimental effects on ecosystems. This study aimed at validating the use of machine learning for recognizing motor anomalies, produced by minimal insecticide exposure on a model insect species. The Mediterranean fruit fly, Ceratitis capitata (Diptera: Tephritidae), was exposed to food contaminated with low concentrations of Carlina acaulis essential oil (EO). A deep learning approach enabled fly pose estimation on video recordings in a custom-built arena. Five machine learning algorithms were trained on handcrafted features, extracted from the predicted pose, to distinguish treated individuals. Random Forest and K-Nearest Neighbor algorithms best performed, with an area under the receiver operating characteristic (ROC) curve of 0.75 and 0.73, respectively. Both algorithms achieved an accuracy of 0.71. Results show the machine learning potential for detecting sublethal effects arising from insecticide exposure on fly motor behavior, which could also affect other organisms and environmental health.

Efficient organisation of the contralateral hemisphere connectome is associated with improvement in intelligence quotient after paediatric epilepsy surgery

OBJECTIVE

Aims of epilepsy surgery in childhood include optimising seizure control and facilitating cognitive development. Predicting which children will improve cognitively is challenging. We investigated the association of the pre-operative structural connectome of the contralateral non-operated hemisphere with improvement in intelligence quotient (IQ) post-operatively.

METHODS

Consecutive children who had undergone unilateral resective procedures for epilepsy at a single centre were retrospectively identified. We included those with pre-operative volume T1-weighted non-contrast brain magnetic resonance imaging (MRI), no visible contralateral MRI abnormalities, and both pre-operative and two years post-operative IQ assessment. The MRI of the hemisphere contralateral to the side of resection was anatomically parcellated into 34 cortical regions and the covariance of cortical thickness between regions was used to create binary and weighted group connectomes.

RESULTS

Eleven patients with a post-operative IQ increase of at least 10 points at two years were compared with twenty-four patients with no change in IQ score. Children who gained at least 10 IQ points post-operatively had a more efficiently structured contralateral hemisphere connectome with higher global efficiency (0.74) compared to those whose IQ did not change at two years (0.58, p = 0.014). This was consistent across thresholds and both binary and weighted networks. There were no statistically significant group differences in age, sex, age at onset of epilepsy, pre-operative IQ, mean cortical thickness, side or site of procedure, two year post-operative Engel scores or use of anti-seizure medications between the two groups.

CONCLUSIONS

Surgical procedures to reduce or stop seizures may allow children with an efficiently structured contralateral hemisphere to achieve their cognitive potential.

Deep learning distinguishes connectomes from focal epilepsy patients and controls: feasibility and clinical implications

The application of deep learning models to evaluate connectome data is gaining interest in epilepsy research. Deep learning may be a useful initial tool to partition connectome data into network subsets for further analysis. Few prior works have used deep learning to examine structural connectomes from patients with focal epilepsy. We evaluated whether a deep learning model applied to whole-brain connectomes could classify 28 participants with focal epilepsy from 20 controls and identify nodal importance for each group. Participants with epilepsy were further grouped based on whether they had focal seizures that evolved into bilateral tonic-clonic seizures (17 with, 11 without). The trained neural network classified patients from controls with an accuracy of 72.92%, while the seizure subtype groups achieved a classification accuracy of 67.86%. In the patient subgroups, the nodes and edges deemed important for accurate classification were also clinically relevant, indicating the model's interpretability. The current work expands the evidence for the potential of deep learning to extract relevant markers from clinical datasets. Our findings offer a rationale for further research interrogating structural connectomes to obtain features that can be biomarkers and aid the diagnosis of seizure subtypes.

Immediate neural impact and incomplete compensation after semantic hub disconnection

The human brain extracts meaning using an extensive neural system for semantic knowledge. Whether broadly distributed systems depend on or can compensate after losing a highly interconnected hub is controversial. We report intracranial recordings from two patients during a speech prediction task, obtained minutes before and after neurosurgical treatment requiring disconnection of the left anterior temporal lobe (ATL), a candidate semantic knowledge hub. Informed by modern diaschisis and predictive coding frameworks, we tested hypotheses ranging from solely neural network disruption to complete compensation by the indirectly affected language-related and speech-processing sites. Immediately after ATL disconnection, we observed neurophysiological alterations in the recorded frontal and auditory sites, providing direct evidence for the importance of the ATL as a semantic hub. We also obtained evidence for rapid, albeit incomplete, attempts at neural network compensation, with neural impact largely in the forms stipulated by the predictive coding framework, in specificity, and the modern diaschisis framework, more generally. The overall results validate these frameworks and reveal an immediate impact and capability of the human brain to adjust after losing a brain hub.

Predicting seizure outcome after epilepsy surgery: Do we need more complex models, larger samples, or better data?

OBJECTIVE

The accurate prediction of seizure freedom after epilepsy surgery remains challenging. We investigated if (1) training more complex models, (2) recruiting larger sample sizes, or (3) using data-driven selection of clinical predictors would improve our ability to predict postoperative seizure outcome using clinical features. We also conducted the first substantial external validation of a machine learning model trained to predict postoperative seizure outcome.

METHODS

We performed a retrospective cohort study of 797 children who had undergone resective or disconnective epilepsy surgery at a tertiary center. We extracted patient information from medical records and trained three models-a logistic regression, a multilayer perceptron, and an XGBoost model-to predict 1-year postoperative seizure outcome on our data set. We evaluated the performance of a recently published XGBoost model on the same patients. We further investigated the impact of sample size on model performance, using learning curve analysis to estimate performance at samples up to N = 2000. Finally, we examined the impact of predictor selection on model performance.

RESULTS

Our logistic regression achieved an accuracy of 72% (95% confidence interval [CI] = 68%-75%, area under the curve [AUC] = .72), whereas our multilayer perceptron and XGBoost both achieved accuracies of 71% (95% CIMLP  = 67%-74%, AUCMLP  = .70; 95% CIXGBoost own  = 68%-75%, AUCXGBoost own  = .70). There was no significant difference in performance between our three models (all p > .4) and they all performed better than the external XGBoost, which achieved an accuracy of 63% (95% CI = 59%-67%, AUC = .62; pLR  = .005, pMLP  = .01, pXGBoost own  = .01) on our data. All models showed improved performance with increasing sample size, but limited improvements beyond our current sample. The best model performance was achieved with data-driven feature selection.

SIGNIFICANCE

We show that neither the deployment of complex machine learning models nor the assembly of thousands of patients alone is likely to generate significant improvements in our ability to predict postoperative seizure freedom. We instead propose that improved feature selection alongside collaboration, data standardization, and model sharing is required to advance the field.

Artificial Intelligence in Neurosurgery: A State-of-the-Art Review from Past to Future

In recent years, there has been a significant surge in discussions surrounding artificial intelligence (AI), along with a corresponding increase in its practical applications in various facets of everyday life, including the medical industry. Notably, even in the highly specialized realm of neurosurgery, AI has been utilized for differential diagnosis, pre-operative evaluation, and improving surgical precision. Many of these applications have begun to mitigate risks of intraoperative and postoperative complications and post-operative care. This article aims to present an overview of the principal published papers on the significant themes of tumor, spine, epilepsy, and vascular issues, wherein AI has been applied to assess its potential applications within neurosurgery. The method involved identifying high-cited seminal papers using PubMed and Google Scholar, conducting a comprehensive review of various study types, and summarizing machine learning applications to enhance understanding among clinicians for future utilization. Recent studies demonstrate that machine learning (ML) holds significant potential in neuro-oncological care, spine surgery, epilepsy management, and other neurosurgical applications. ML techniques have proven effective in tumor identification, surgical outcomes prediction, seizure outcome prediction, aneurysm prediction, and more, highlighting its broad impact and potential in improving patient management and outcomes in neurosurgery. This review will encompass the current state of research, as well as predictions for the future of AI within neurosurgery.

New insights for predicting surgery outcome in patients with temporal lobe epilepsy. A systematic review

Resective surgery is the treatment of choice for one-third of adult patients with focal, drug-resistant epilepsy. This procedure is associated with substantial clinical and cognitive risks. In clinical practice, there is no validated model for epilepsy surgery outcome prediction (ESOP). Meta-analyses on ESOP studies assessing prognostic factors report discrepancies in terms of study design. Our review aims to systematically investigate methodological and analytical aspects of studies predicting clinical and cognitive outcomes after temporal lobe epilepsy surgery. A systematic review of ESOP studies published between 2000 and 2022 from three databases (MEDLINE, Web of Science, and PsycINFO) was completed by following the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines. It yielded 4867 articles. Among them, 21 corresponded to our inclusion criteria and were therefore retained in the final review. The risk of bias was assessed using A Tool to Assess Risk of Bias and Applicability of Prediction Model Studies (PROBAST). Data extracted from the 21 studies were analyzed using narrative synthesis and descriptive statistics. Our findings show an increase in the use of multimodal datasets and machine learning analyses in recent ESOP studies, although regression remained the most frequently used approach. We also identified a more frequent use of network notions in recent ESOP studies. Nevertheless, several methodological issues were noted, such as small sample sizes, lack of information on the follow-up period, variability in seizure outcome, and the definition of neuropsychological postoperative change. Of 21 studies, only one provided a clinical tool to anticipate the cognitive outcome after epilepsy surgery. We conclude that methodological issues should be overcome before we move towards more complete models to better predict clinical and cognitive outcomes after epilepsy surgery. Recommendations for future studies to harness the possibilities of multimodal datasets and data fusion, are provided. A stronger bridge between fundamental and clinical research may result in developing accessible clinical tools.

Pre-surgical features of intrinsic brain networks predict single and joint epilepsy surgery outcomes

Despite the effectiveness of surgical interventions for the treatment of intractable focal temporal lobe epilepsy (TLE), the substrates that support good outcomes are poorly understood. While algorithms have been developed for the prediction of either seizure or cognitive/psychiatric outcomes alone, no study has reported on the functional and structural architecture that supports joint outcomes. We measured key aspects of pre-surgical whole brain functional/structural network architecture and evaluated their ability to predict post-operative seizure control in combination with cognitive/psychiatric outcomes. Pre-surgically, we identified the intrinsic connectivity networks (ICNs) unique to each person through independent component analysis (ICA), and computed: (1) the spatial-temporal match between each person's ICA components and established, canonical ICNs, (2) the connectivity strength within each identified person-specific ICN, (3) the gray matter (GM) volume underlying the person-specific ICNs, and (4) the amount of variance not explained by the canonical ICNs for each person. Post-surgical seizure control and reliable change indices of change (for language [naming, phonemic fluency], verbal episodic memory, and depression) served as binary outcome responses in random forest (RF) models. The above functional and structural measures served as input predictors. Our empirically derived ICN-based measures customized to the individual showed that good joint seizure and cognitive/psychiatric outcomes depended upon higher levels of brain reserve (GM volume) in specific networks. In contrast, singular outcomes relied on systematic, idiosyncratic variance in the case of seizure control, and the weakened pre-surgical presence of functional ICNs that encompassed the ictal temporal lobe in the case of cognitive/psychiatric outcomes. Our data made clear that the ICNs differed in their propensity to provide reserve for adaptive outcomes, with some providing structural (brain), and others functional (cognitive) reserve. Our customized methodology demonstrated that when substantial unique, patient-specific ICNs are present prior to surgery there is a reliable association with poor post-surgical seizure control. These ICNs are idiosyncratic in that they did not match the canonical, normative ICNs and, therefore, could not be defined functionally, with their location likely varying by patient. This important finding suggested the level of highly individualized ICN's in the epileptic brain may signal the emergence of epileptogenic activity after surgery.

Machine Learning for Precision Epilepsy Surgery

Brain surgery offers the best chance of seizure-freedom for patients with focal drug-resistant epilepsy, but only 50% achieve sustained seizure-freedom. With the explosion of data collected during routine presurgical evaluations and recent advances in computational science, we now have a tremendous potential to achieve precision epilepsy surgery: a data-driven tailoring of surgical planning. This review highlights the clinical need, the relevant computational science focusing on machine learning, and discusses some specific applications in epilepsy surgery.

Brain functional activity-based classification of autism spectrum disorder using an attention-based graph neural network combined with gene expression

Autism spectrum disorder (ASD) is a complex brain neurodevelopmental disorder related to brain activity and genetics. Most of the ASD diagnostic models perform feature selection at the group level without considering individualized information. Evidence has shown the unique topology of the individual brain has a fundamental impact on brain diseases. Thus, a data-constructing method fusing individual topological information and a corresponding classification model is crucial in ASD diagnosis and biomarker discovery. In this work, we trained an attention-based graph neural network (GNN) to perform the ASD diagnosis with the fusion of graph data. The results achieved an accuracy of 79.78%. Moreover, we found the model paid high attention to brain regions mainly involved in the social-brain circuit, default-mode network, and sensory perception network. Furthermore, by analyzing the covariation between functional magnetic resonance imaging data and gene expression, current studies detected several ASD-related genes (i.e. MUTYH, AADAT, and MAP2), and further revealed their links to image biomarkers. Our work demonstrated that the ASD diagnostic framework based on graph data and attention-based GNN could be an effective tool for ASD diagnosis. The identified functional features with high attention values may serve as imaging biomarkers for ASD.

Protocol for an observational cohort study investigating biomarkers predicting seizure recurrence following a first unprovoked seizure in adults

INTRODUCTION

A first unprovoked seizure is a common presentation, reliably identifying those that will have recurrent seizures is a challenge. This study will be the first to explore the combined utility of serum biomarkers, quantitative electroencephalogram (EEG) and quantitative MRI to predict seizure recurrence. This will inform patient stratification for counselling and the inclusion of high-risk patients in clinical trials of disease-modifying agents in early epilepsy.

METHODS AND ANALYSIS

100 patients with first unprovoked seizure will be recruited from a tertiary neuroscience centre and baseline assessments will include structural MRI, EEG and a blood sample. As part of a nested pilot study, a subset of 40 patients will have advanced MRI sequences performed that are usually reserved for patients with refractory chronic epilepsy. The remaining 60 patients will have standard clinical MRI sequences. Patients will be followed up every 6 months for a 24-month period to assess seizure recurrence. Connectivity and network-based analyses of EEG and MRI data will be carried out and examined in relation to seizure recurrence. Patient outcomes will also be investigated with respect to analysis of high-mobility group box-1 from blood serum samples.

ETHICS AND DISSEMINATION

This study was approved by North East-Tyne & Wear South Research Ethics Committee (20/NE/0078) and funded by an Association of British Neurologists and Guarantors of Brain clinical research training fellowship. Findings will be presented at national and international meetings published in peer-reviewed journals.

TRIAL REGISTRATION NUMBER

NIHR Clinical Research Network's (CRN) Central Portfolio Management System (CPMS)-44976.

Diffusion Kernel Attention Network for Brain Disorder Classification

Constructing and analyzing functional brain networks (FBN) has become a promising approach to brain disorder classification. However, the conventional successive construct-and-analyze process would limit the performance due to the lack of interactions and adaptivity among the subtasks in the process. Recently, Transformer has demonstrated remarkable performance in various tasks, attributing to its effective attention mechanism in modeling complex feature relationships. In this paper, for the first time, we develop Transformer for integrated FBN modeling, analysis and brain disorder classification with rs-fMRI data by proposing a Diffusion Kernel Attention Network to address the specific challenges. Specifically, directly applying Transformer does not necessarily admit optimal performance in this task due to its extensive parameters in the attention module against the limited training samples usually available. Looking into this issue, we propose to use kernel attention to replace the original dot-product attention module in Transformer. This significantly reduces the number of parameters to train and thus alleviates the issue of small sample while introducing a non-linear attention mechanism to model complex functional connections. Another limit of Transformer for FBN applications is that it only considers pair-wise interactions between directly connected brain regions but ignores the important indirect connections. Therefore, we further explore diffusion process over the kernel attention to incorporate wider interactions among indirectly connected brain regions. Extensive experimental study is conducted on ADHD-200 data set for ADHD classification and on ADNI data set for Alzheimer's disease classification, and the results demonstrate the superior performance of the proposed method over the competing methods.

Altered Structural Brain Networks in Refractory and Nonrefractory Idiopathic Generalized Epilepsy

Introduction: Idiopathic generalized epilepsy (IGE) is a collection of generalized nonlesional epileptic network disorders. Around 20-40% of patients with IGE are refractory to antiseizure medication, and mechanisms underlying refractoriness are poorly understood. Here, we characterize structural brain network alterations and determine whether network alterations differ between patients with refractory and nonrefractory IGE. Methods: Thirty-three patients with IGE (10 nonrefractory and 23 refractory) and 39 age- and sex-matched healthy controls were studied. Network nodes were segmented from T1-weighted images, while connections between these nodes (edges) were reconstructed from diffusion magnetic resonance imaging (MRI). Diffusion networks of fractional anisotropy (FA), mean diffusivity (MD), radial diffusivity (RD), and streamline count (Count) were studied. Differences between all patients, refractory, nonrefractory, and control groups were computed using network-based statistics. Nodal volume differences between groups were computed using Cohen's d effect size calculation. Results: Patients had significantly decreased bihemispheric FA and Count networks and increased MD and RD networks compared with controls. Alterations in network architecture, with respect to controls, differed depending on treatment outcome, including predominant FA network alterations in refractory IGE and increased nodal volume in nonrefractory IGE. Diffusion MRI networks were not influenced by nodal volume. Discussion: Although a nonlesional disorder, patients with IGE have bihemispheric structural network alterations that may differ between patients with refractory and nonrefractory IGE. Given that distinct nodal volume and FA network alterations were observed between treatment outcome groups, a multifaceted network analysis may be useful for identifying imaging biomarkers of refractory IGE.

Volumetric and structural connectivity abnormalities co-localise in TLE

Patients with temporal lobe epilepsy (TLE) exhibit both volumetric and structural connectivity abnormalities relative to healthy controls. How these abnormalities inter-relate and their mechanisms are unclear. We computed grey matter volumetric changes and white matter structural connectivity abnormalities in 144 patients with unilateral TLE and 96 healthy controls. Regional volumes were calculated using T1-weighted MRI, while structural connectivity was derived using white matter fibre tractography from diffusion-weighted MRI. For each regional volume and each connection strength, we calculated the effect size between patient and control groups in a group-level analysis. We then applied hierarchical regression to investigate the relationship between volumetric and structural connectivity abnormalities in individuals. Additionally, we quantified whether abnormalities co-localised within individual patients by computing Dice similarity scores. In TLE, white matter connectivity abnormalities were greater when joining two grey matter regions with abnormal volumes. Similarly, grey matter volumetric abnormalities were greater when joined by abnormal white matter connections. The extent of volumetric and connectivity abnormalities related to epilepsy duration, but co-localisation did not. Co-localisation was primarily driven by neighbouring abnormalities in the ipsilateral hemisphere. Overall, volumetric and structural connectivity abnormalities were related in TLE. Our results suggest that shared mechanisms may underlie changes in both volume and connectivity alterations in patients with TLE.

Altered Structural and Functional Patterns Within Executive Control Network Distinguish Frontal Glioma-Related Epilepsy

Background

The tumor invasion of the frontal lobe induces changes in the executive control network (ECN). It remains unclear whether epileptic seizures in frontal glioma patients exacerbate the structural and functional alterations within the ECN, and whether these changes can be used to identify glioma-related seizures at an early stage. This study aimed to investigate the altered structural and functional patterns of ECN in frontal gliomas without epilepsy (non-FGep) and frontal gliomas with epilepsy (FGep) and to evaluate whether the patterns can accurately distinguish glioma-related epilepsy.

Methods

We measured gray matter (GM) volume, regional homogeneity (ReHo), and functional connectivity (FC) within the ECN to identify the structural and functional changes in 50 patients with frontal gliomas (29 non-FGep and 21 FGep) and 39 healthy controls (CN). We assessed the relationships between the structural and functional changes and cognitive function using partial correlation analysis. Finally, we applied a pattern classification approach to test whether structural and functional abnormalities within the ECN can distinguish non-FGep and FGep from CN subjects.

Results

Within the ECN, non-FGep and FGep showed increased local structure (GM) and function (ReHo), and decreased FC between brain regions compared to CN. Also, non-FGep and FGep showed differential patterns of structural and functional abnormalities within the ECN, and these abnormalities are more severe in FGep than in non-FGep. Lastly, FC between the right superior frontal gyrus and right dorsolateral prefrontal cortex was positively correlated with episodic memory scores in non-FGep and FGep. In particular, the support vector machine (SVM) classifier based on structural and functional abnormalities within ECN could accurately distinguish non-FGep and FGep from CN, and FGep from non-FGep on an individual basis with very high accuracy, area under the curve (AUC), sensitivity, and specificity.

Conclusion

Tumor invasion of the frontal lobe induces local structural and functional reorganization within the ECN, exacerbated by the accompanying epileptic seizures. The ECN abnormalities can accurately distinguish the presence or absence of epileptic seizures in frontal glioma patients. These findings suggest that differential ECN patterns can assist in the early identification and intervention of epileptic seizures in frontal glioma patients.

Normative brain mapping of interictal intracranial EEG to localize epileptogenic tissue

The identification of abnormal electrographic activity is important in a wide range of neurological disorders, including epilepsy for localizing epileptogenic tissue. However, this identification may be challenging during non-seizure (interictal) periods, especially if abnormalities are subtle compared to the repertoire of possible healthy brain dynamics. Here, we investigate if such interictal abnormalities become more salient by quantitatively accounting for the range of healthy brain dynamics in a location-specific manner. To this end, we constructed a normative map of brain dynamics, in terms of relative band power, from interictal intracranial recordings from 234 participants (21 598 electrode contacts). We then compared interictal recordings from 62 patients with epilepsy to the normative map to identify abnormal regions. We proposed that if the most abnormal regions were spared by surgery, then patients would be more likely to experience continued seizures postoperatively. We first confirmed that the spatial variations of band power in the normative map across brain regions were consistent with healthy variations reported in the literature. Second, when accounting for the normative variations, regions that were spared by surgery were more abnormal than those resected only in patients with persistent postoperative seizures (t = -3.6, P = 0.0003), confirming our hypothesis. Third, we found that this effect discriminated patient outcomes (area under curve 0.75 P = 0.0003). Normative mapping is a well-established practice in neuroscientific research. Our study suggests that this approach is feasible to detect interictal abnormalities in intracranial EEG, and of potential clinical value to identify pathological tissue in epilepsy. Finally, we make our normative intracranial map publicly available to facilitate future investigations in epilepsy and beyond.

Neuroimaging gradient alterations and epileptogenic prediction in focal cortical dysplasia Ⅲa

INTRODUCTION

Focal cortical dysplasia Type Ⅲa (FCD Ⅲa) is a highly prevalent temporal lobe epilepsy but the seizure outcomes are not satisfactory after epilepsy surgery. Hence, quantitative neuroimaging, epileptogenic alterations, as well as their values in guiding surgery are worth exploring.

METHODS

We examined 69 patients with pathologically verified FCD Ⅲa using multimodal neuroimaging and stereoelectroencephalography (SEEG). Among them, 18 received postoperative imaging which showed the extent of surgical resection and 9 underwent SEEG implantation. We also explored neuroimaging gradient alterations along with the distance to the temporal pole. Subsequently, the machine learning regression model was employed to predict whole-brain epileptogenicity. Lastly, the correlation between neuroimaging or epileptogenicity and surgical cavities was assessed.

RESULTS

FCD Ⅲa displayed neuroimaging gradient alterations on the temporal neocortex, morphology-signal intensity decoupling, low similarity of intra-morphological features and high similarity of intra-signal intensity features. The support vector regression model was successfully applied at the whole-brain level to calculate the continuous epileptogenic value at each vertex (mean-squared error = 13.8 ± 9.8).

CONCLUSION

Our study investigated the neuroimaging gradient alterations and epileptogenicity of FCD Ⅲa, along with their potential values in guiding suitable resection range and in predicting postoperative seizure outcomes. The conclusions from this study may facilitate an accurate presurgical examination of FCD Ⅲa. However, further investigation including a larger cohort is necessary to confirm the results.

Deep learning derived automated ASPECTS on non-contrast CT scans of acute ischemic stroke patients

Ischemic stroke is the most common type of stroke, ranked as the second leading cause of death worldwide. The Alberta Stroke Program Early CT Score (ASPECTS) is considered as a systematic method of assessing ischemic change on non-contrast CT scans (NCCT) of acute ischemic stroke (AIS) patients, while still suffering from the requirement of experts' experience and also the inconsistent results between readers. In this study, we proposed an automated ASPECTS method to utilize the powerful learning ability of neural networks for objectively scoring CT scans of AIS patients. First, we proposed to use the CT perfusion (CTP) from one-stop stroke imaging to provide the golden standard of ischemic regions for ASPECTS scoring. Second, we designed an asymmetry network to capture features when comparing the left and right sides for each ASPECTS region to estimate its ischemic status. Third, we performed experiments in a large main dataset of 870 patients, as well as an independent testing dataset consisting of 207 patients with radiologists' scorings. Experimental results show that our network achieved remarkable performance, as sensitivity and accuracy of 93.7 and 92.4% in the main dataset, and 95.5 and 91.3% in the independent testing dataset, respectively. In the latter dataset, our analysis revealed a high positive correlation between the ASPECTS score and the prognosis of patients in 90DmRs. Also, we found ASPECTS score is a good indicator of the size of CTP core volume of an infraction. The proposed method shows its potential for automated ASPECTS scoring on NCCT images.

Individual Brain Metabolic Connectome Indicator Based on Jensen-Shannon Divergence Similarity Estimation Predicts Seizure Outcomes of Temporal Lobe Epilepsy

Objective: We aimed to use an individual metabolic connectome method, the Jensen-Shannon Divergence Similarity Estimation (JSSE), to characterize the aberrant connectivity patterns and topological alterations of the individual-level brain metabolic connectome and predict the long-term surgical outcomes in temporal lobe epilepsy (TLE).

Methods: A total of 128 patients with TLE (63 females, 65 males; 25.07 ± 12.01 years) who underwent Positron emission tomography (PET) with 18F-fluorodeoxyglucose (FDG) imaging were enrolled. Patients were classified either as experiencing seizure recurrence (SZR) or seizure free (SZF) at least 1 year after surgery. Each individual’s metabolic brain network was ascertained using the proposed JSSE method. We compared the similarity and difference in the JSSE network and its topological measurements between the two groups. The two groups were then classified by combining the information from connection and topological metrics, which was conducted by the multiple kernel support vector machine. The validation was performed using the nested leave-one-out cross-validation strategy to confirm the performance of the methods.

Results: With a median follow-up of 33 months, 50% of patients achieved SZF. No relevant differences in clinical features were found between the two groups except age at onset. The proposed JSSE method showed marked degree reductions in IFGoperc.R, ROL. R, IPL. R, and SMG. R; and betweenness reductions in ORBsup.R and IOG. R; meanwhile, it found increases in the degree analysis of CAL. L and PCL. L, and in the betweenness analysis of PreCG.R, IOG. R, PoCG.R, PCL. L and PCL.R. Exploring consensus significant metabolic connections, we observed that the most involved metabolic motor networks were the INS-TPOmid.L, MTG. R-SMG. R, and MTG. R-IPL.R pathways between the two groups, and yielded another detailed individual pathological connectivity in the PHG. R-CAU.L, PHG. R-HIP.L, TPOmid.L-LING.R, TPOmid.L-DCG.R, MOG. R-MTG.R, MOG. R-ANG.R, and IPL. R-IFGoperc.L pathways. These aberrant functional network measures exhibited ideal classification performance in predicting SZF individuals from SZR ones at a sensitivity of 75.00%, a specificity of 92.79%, and an accuracy of 83.59%.

Conclusion: The JSSE method indicator can identify abnormal brain networks in predicting an individual’s long-term surgical outcome of TLE, thus potentially constituting a clinically applicable imaging biomarker. The results highlight the biological meaning of the estimated individual brain metabolic connectome.

Machine Learning Approaches for Imaging-Based Prognostication of the Outcome of Surgery for Mesial Temporal Lobe Epilepsy

Objectives

Around 30% of patients undergoing surgical resection for drug‐resistant mesial temporal lobe epilepsy (MTLE) do not obtain seizure freedom. Success of anterior temporal lobe resection (ATLR) critically depends on the careful selection of surgical candidates, aiming at optimizing seizure freedom while minimizing postoperative morbidity. Structural MRI and FDG‐PET neuroimaging are routinely used in presurgical assessment and guide the decision to proceed to surgery. In this study, we evaluate the potential of machine learning techniques applied to standard presurgical MRI and PET imaging features to provide enhanced prognostic value relative to current practice.

Methods

Eighty two patients with drug resistant MTLE were scanned with FDG‐PET pre‐surgery and T1‐weighted MRI pre‐ and postsurgery. From these images the following features of interest were derived: volume of temporal lobe (TL) hypometabolism, % of extratemporal hypometabolism, presence of contralateral TL hypometabolism, presence of hippocampal sclerosis, laterality of seizure onset volume of tissue resected and % of temporal lobe hypometabolism resected. These measures were used as predictor variables in logistic regression, support vector machines, random forests and artificial neural networks.

Results

In the study cohort, 24 of 82 (28.3%) who underwent an ATLR for drug‐resistant MTLE did not achieve Engel Class I (i.e., free of disabling seizures) outcome at a minimum of 2 years of postoperative follow‐up. We found that machine learning approaches were able to predict up to 73% of the 24 ATLR surgical patients who did not achieve a Class I outcome, at the expense of incorrect prediction for up to 31% of patients who did achieve a Class I outcome. Overall accuracies ranged from 70% to 80%, with an area under the receiver operating characteristic curve (AUC) of .75–.81. We additionally found that information regarding overall extent of both total and significantly hypometabolic tissue resected was crucial to predictive performance, with AUC dropping to .59–.62 using presurgical information alone. Incorporating the laterality of seizure onset and the choice of machine learning algorithm did not significantly change predictive performance.

Significance

Collectively, these results indicate that "acceptable" to "good" patient‐specific prognostication for drug‐resistant MTLE surgery is feasible with machine learning approaches utilizing commonly collected imaging modalities, but that information on the surgical resection region is critical for optimal prognostication.

ConCeptCNN: A novel multi-filter convolutional neural network for the prediction of neurodevelopmental disorders using brain connectome

BACKGROUND

Deep convolutional neural network (CNN) and its derivatives have recently shown great promise in the prediction of brain disorders using brain connectome data. Existing deep CNN methods using single global row and column convolutional filters have limited ability to extract discriminative information from brain connectome for prediction tasks.

PURPOSE

This paper presents a novel deep Connectome-Inception CNN (ConCeptCNN) model, which is developed based on multiple convolutional filters. The proposed model is used to extract topological features from brain connectome data for neurological disorders classification and analysis.

METHODS

The ConCeptCNN uses multiple vector-shaped filters extract topological information from the brain connectome at different levels for complementary feature embeddings of brain connectome. The proposed model is validated using two datasets: the Neuro Bureau ADHD-200 dataset and the Cincinnati Early Prediction Study (CINEPS) dataset.

RESULTS

In a cross-validation experiment, the ConCeptCNN achieved a prediction accuracy of 78.7% for the detection of ADHD in adolescents and an accuracy of 81.6% for the prediction of cognitive deficits at 2-year corrected age in very preterm infants. In addition to the classification tasks, the ConCeptCNN identified several brain regions that are discriminative to neurodevelopmental disorders.

CONCLUSIONS

We compared the ConCeptCNN with several peer CNN methods. The results demonstrated that proposed model improves overall classification performance of neurodevelopmental disorders prediction tasks. This article is protected by copyright. All rights reserved.

A hybrid deep learning model for forecasting lymphocyte depletion during radiation therapy

PURPOSE

Recent studies have shown that severe depletion of the absolute lymphocyte count (ALC) induced by radiation therapy (RT) has been associated with poor overall survival of patients with many solid tumors. In this paper, we aimed to predict radiation-induced lymphocyte depletion in esophageal cancer patients during the course of RT based on patient characteristics and dosimetric features.

METHODS

We proposed a hybrid deep learning model in a stacked structure to predict a trend toward ALC depletion based on the clinical information before or at the early stages of RT treatment. The proposed model consisted of four channels, one channel based on long short-term memory (LSTM) network and three channels based on neural networks, to process four categories of features followed by a dense layer to integrate the outputs of four channels and predict the weekly ALC values. Moreover, a discriminative kernel was developed to extract temporal features and assign different weights to each part of the input sequence which enabled the model to focus on the most relevant parts. The proposed model was trained and tested on a dataset of 860 esophageal cancer patients who received concurrent chemoradiotherapy.

RESULTS

The performance of the proposed model was evaluated based on several important prediction metrics and compared to other commonly used prediction models. The results showed that the proposed model outperformed off-the-shelf prediction methods with at least a 30% reduction in the mean squared error (MSE) of weekly ALC predictions based on pretreatment data.Moreover, using an extended model based on augmented first-week treatment data reduced the MSE of predictions by 70% compared to the model based on the pretreatment data.

CONCLUSIONS

In conclusion, our model performed well in predicting radiation-induced lymphocyte depletion for RT treatment planning. The ability to predict ALC will enable physicians to evaluate individual RT treatment plans for lymphopenia risk and to identify patients at high risk who would benefit from modified treatment approaches. This article is protected by copyright. All rights reserved.

Machine learning applications on neuroimaging for diagnosis and prognosis of epilepsy: A review

Machine learning is playing an increasingly important role in medical image analysis, spawning new advances in the clinical application of neuroimaging. There have been some reviews on machine learning and epilepsy before, and they mainly focused on electrophysiological signals such as electroencephalography (EEG) and stereo electroencephalography (SEEG), while neglecting the potential of neuroimaging in epilepsy research. Neuroimaging has its important advantages in confirming the range of the epileptic region, which is essential in presurgical evaluation and assessment after surgery. However, it is difficult for EEG to locate the accurate epilepsy lesion region in the brain. In this review, we emphasize the interaction between neuroimaging and machine learning in the context of epilepsy diagnosis and prognosis. We start with an overview of epilepsy and typical neuroimaging modalities used in epilepsy clinics, MRI, DWI, fMRI, and PET. Then, we elaborate two approaches in applying machine learning methods to neuroimaging data: (i) the conventional machine learning approach combining manual feature engineering and classifiers, (ii) the deep learning approach, such as the convolutional neural networks and autoencoders. Subsequently, the application of machine learning on epilepsy neuroimaging, such as segmentation, localization, and lateralization tasks, as well as tasks directly related to diagnosis and prognosis are looked into in detail. Finally, we discuss the current achievements, challenges, and potential future directions in this field, hoping to pave the way for computer-aided diagnosis and prognosis of epilepsy.

Machine Learning in Neuro-Oncology, Epilepsy, Alzheimer's Disease, and Schizophrenia

Applications of machine learning (ML) in translational medicine include therapeutic drug creation, diagnostic development, surgical planning, outcome prediction, and intraoperative assistance. Opportunities in the neurosciences are rich given advancement in our understanding of the brain, expanding indications for intervention, and diagnostic challenges often characterized by multiple clinical and environmental factors. We present a review of ML in neuro-oncology, epilepsy, Alzheimer's disease, and schizophrenia to highlight recent progression in these field, optimizing machine learning capabilities in their current forms. Supervised learning models appear to be the most commonly incorporated algorithm models for machine learning across the reviewed neuroscience disciplines with primary aim of diagnosis. Accuracy ranges are high from 63% to 99% across all algorithms investigated. Machine learning contributions to neurosurgery, neurology, psychiatry, and the clinical and basic science neurosciences may enhance current medical best practices while also broadening our understanding of dynamic neural networks and the brain.

A Brief History of Machine Learning in Neurosurgery

The history of machine learning in neurosurgery spans three decades and continues to develop at a rapid pace. The earliest applications of machine learning within neurosurgery were first published in the 1990s as researchers began developing artificial neural networks to analyze structured datasets and supervised tasks. By the turn of the millennium, machine learning had evolved beyond proof-of-concept; algorithms had success detecting tumors in unstructured clinical imaging, and unsupervised learning showed promise for tumor segmentation. Throughout the 2000s, the role of machine learning in neurosurgery was further refined. Well-trained models began to consistently best expert clinicians at brain tumor diagnosis. Additionally, the digitization of the healthcare industry provided ample data for analysis, both structured and unstructured. By the 2010s, the use of machine learning within neurosurgery had exploded. The rapid deployment of an exciting new toolset also led to the growing realization that it may offer marginal benefit at best over conventional logistical regression models for analyzing tabular datasets. Additionally, the widespread adoption of machine learning in neurosurgical clinical practice continues to lag until additional validation can ensure generalizability. Many exciting contemporary applications nonetheless continue to demonstrate the unprecedented potential of machine learning to revolutionize neurosurgery when applied to appropriate clinical challenges.

Effects of laser interstitial thermal therapy for mesial temporal lobe epilepsy on the structural connectome and its relationship to seizure freedom

OBJECTIVE

Laser interstitial thermal therapy (LITT) is a minimally invasive surgery for mesial temporal lobe epilepsy (mTLE), but the effects of individual patient anatomy and location of ablation volumes affect seizure outcomes. The purpose of this study is to see if features of individual patient structural connectomes predict surgical outcomes after LITT for mTLE.

METHODS

This is a retrospective analysis of seizure outcomes of LITT for mTLE in 24 patients. We use preoperative diffusion tensor imaging (DTI) to simulate changes in structural connectivity after laser ablation. A two-step machine-learning algorithm is applied to predict seizure outcomes from the change in connectomic features after surgery.

RESULTS

Although node-based network features such as clustering coefficient and betweenness centrality have some predictive value, changes in connection strength between mesial temporal regions predict seizure outcomes significantly better. Changes in connection strength between the entorhinal cortex (EC), and the insula, hippocampus, and amygdala, as well as between the temporal pole and hippocampus, predict Engel Class I outcomes with an accuracy of 88%. Analysis of the ablation location, as well as simulated, alternative ablations, reveals that a more medial, anterior, and inferior ablation volume is associated with a greater effect on these connections, and potentially on seizure outcomes.

SIGNIFICANCE

Our results indicate (1) that seizure outcomes can be retrospectively predicted with excellent accuracy using changes in structural connectivity, and (2) that favorable connectomic changes are associated with an ablation volume involving relatively mesial, anterior, and inferior locations. These results may provide a framework whereby individual pre-operative structural connectomes can be used to optimize ablation volumes and improve outcomes in LITT for mTLE.

Can we predict anti-seizure medication response in focal epilepsy using machine learning?

OBJECTIVE

The aim of this study was to evaluate the feasibility of machine learning approach based on clinical factors and diffusion tensor imaging (DTI) to predict anti-seizure medication (ASM) response in focal epilepsy. We hypothesized that ASM response in focal epilepsy can be predicted using a machine learning approach.

METHODS

In this retrospective study conducted at a tertiary hospital, we enrolled 160 patients with newly diagnosed focal epilepsy. All of them underwent DTI from January 2017 to July 2019, with a follow-up at least 12 months after the diagnosis of epilepsy based on regular evaluation of ASMs. We analyzed the patients' clinical characteristics, and the conventional DTI measurements and extracted the structural connectomic profiles from the DTI. We employed the support vector machine (SVM) algorithm, and a k-fold cross-validation was executed.

RESULTS

The highest accuracy of classification was ensured based on the clinical factors. A SVM classifier based on the clinical factors revealed an accuracy of 87.5% and an area under curve (AUC) of 0.882. Another SVM classifier based on the conventional DTI measures demonstrated an accuracy of 62.5% and an AUC of 0.611. In addition, an SVM classifier based on the structural connectomic profiles revealed an accuracy of 68.7% and an AUC of 0.667. The AUC of the ROC curve generated from the clinical factors was significantly higher than the ROC curves based on the conventional DTI measures or structural connectomic profiles.

CONCLUSION

Machine learning approach is useful in predicting the ASM response in focal epilepsy. The clinical factor is more important than the conventional DTI measures and structural connectomic profiles in predicting the ASM response in focal epilepsy.

White matter structural connectivity as a biomarker for detecting juvenile myoclonic epilepsy by transferred deep convolutional neural networks with varying transfer rates

Objective. By detecting abnormal white matter changes, diffusion magnetic resonance imaging (MRI) contributes to the detection of juvenile myoclonic epilepsy (JME). In addition, deep learning has greatly improved the detection performance of various brain disorders. However, there is almost no previous study effectively detecting JME by a deep learning approach with diffusion MRI.Approach. In this study, the white matter structural connectivity was generated by tracking the white matter fibers in detail based on Q-ball imaging and neurite orientation dispersion and density imaging. Four advanced deep convolutional neural networks (CNNs) were deployed by using the transfer learning approach, in which the transfer rate searching strategy was proposed to achieve the best detection performance.Main results. Our results showed: (a) Compared to normal control, the white matter' neurite density of JME was significantly decreased. The most significantly abnormal fiber tracts between the two groups were found to be cortico-cortical connection tracts. (b) The proposed transfer rate searching approach contributed to find each CNN's best performance, in which the best JME detection accuracy of 92.2% was achieved by using the Inception_resnet_v2 network with a 16% transfer rate.Significance. The results revealed: (a) Through detection of the abnormal white matter changes, the white matter structural connectivity can be used as a useful biomarker for detecting JME, which helps to characterize the pathophysiology of epilepsy. (b) The proposed transfer rate, as a new hyperparameter, promotes the CNNs transfer learning performance in detecting JME.

Machine learning models for decision support in epilepsy management: A critical review

PURPOSE

There remain major challenges for the clinician in managing patients with epilepsy effectively. Choosing anti-seizure medications (ASMs) is subject to trial and error. About one-third of patients have drug-resistant epilepsy (DRE). Surgery may be considered for selected patients, but time from diagnosis to surgery averages 20 years. We reviewed the potential use of machine learning (ML) predictive models as clinical decision support tools to help address some of these issues.

METHODS

We conducted a comprehensive search of Medline and Embase of studies that investigated the application of ML in epilepsy management in terms of predicting ASM responsiveness, predicting DRE, identifying surgical candidates, and predicting epilepsy surgery outcomes. Original articles addressing these 4 areas published in English between 2000 and 2020 were included.

RESULTS

We identified 24 relevant articles: 6 on ASM responsiveness, 3 on DRE prediction, 2 on identifying surgical candidates, and 13 on predicting surgical outcomes. A variety of potential predictors were used including clinical, neuropsychological, imaging, electroencephalography, and health system claims data. A number of different ML algorithms and approaches were used for prediction, but only one study utilized deep learning methods. Some models show promising performance with areas under the curve above 0.9. However, most were single setting studies (18 of 24) with small sample sizes (median number of patients 55), with the exception of 3 studies that utilized large databases and 3 studies that performed external validation. There was a lack of standardization in reporting model performance. None of the models reviewed have been prospectively evaluated for their clinical benefits.

CONCLUSION

The utility of ML models for clinical decision support in epilepsy management remains to be determined. Future research should be directed toward conducting larger studies with external validation, standardization of reporting, and prospective evaluation of the ML model on patient outcomes.

The Rise of Wearable Devices during the COVID-19 Pandemic: A Systematic Review

The COVID-19 pandemic has wreaked havoc globally and still persists even after a year of its initial outbreak. Several reasons can be considered: people are in close contact with each other, i.e., at a short range (1 m), and the healthcare system is not sufficiently developed or does not have enough facilities to manage and fight the pandemic, even in developed countries such as the USA and the U.K. and countries in Europe. There is a great need in healthcare for remote monitoring of COVID-19 symptoms. In the past year, a number of IoT-based devices and wearables have been introduced by researchers, providing good results in terms of high accuracy in diagnosing patients in the prodromal phase and in monitoring the symptoms of patients, i.e., respiratory rate, heart rate, temperature, etc. In this systematic review, we analyzed these wearables and their need in the healthcare system. The research was conducted using three databases: IEEE Xplore®, Web of Science®, and PubMed Central®, between December 2019 and June 2021. This article was based on the PRISMA guidelines. Initially, 1100 articles were identified while searching the scientific literature regarding this topic. After screening, ultimately, 70 articles were fully evaluated and included in this review. These articles were divided into two categories. The first one belongs to the on-body sensors (wearables), their types and positions, and the use of AI technology with ehealth wearables in different scenarios from screening to contact tracing. In the second category, we discuss the problems and solutions with respect to utilizing these wearables globally. This systematic review provides an extensive overview of wearable systems for the remote management and automated assessment of COVID-19, taking into account the reliability and acceptability of the implemented technologies.

Artificial intelligence for classification of temporal lobe epilepsy with ROI-level MRI data: A worldwide ENIGMA-Epilepsy study

Artificial intelligence has recently gained popularity across different medical fields to aid in the detection of diseases based on pathology samples or medical imaging findings. Brain magnetic resonance imaging (MRI) is a key assessment tool for patients with temporal lobe epilepsy (TLE). The role of machine learning and artificial intelligence to increase detection of brain abnormalities in TLE remains inconclusive. We used support vector machine (SV) and deep learning (DL) models based on region of interest (ROI-based) structural (n = 336) and diffusion (n = 863) brain MRI data from patients with TLE with ("lesional") and without ("non-lesional") radiographic features suggestive of underlying hippocampal sclerosis from the multinational (multi-center) ENIGMA-Epilepsy consortium. Our data showed that models to identify TLE performed better or similar (68-75%) compared to models to lateralize the side of TLE (56-73%, except structural-based) based on diffusion data with the opposite pattern seen for structural data (67-75% to diagnose vs. 83% to lateralize). In other aspects, structural and diffusion-based models showed similar classification accuracies. Our classification models for patients with hippocampal sclerosis were more accurate (68-76%) than models that stratified non-lesional patients (53-62%). Overall, SV and DL models performed similarly with several instances in which SV mildly outperformed DL. We discuss the relative performance of these models with ROI-level data and the implications for future applications of machine learning and artificial intelligence in epilepsy care.

Incorporation of quantitative MRI in a model to predict temporal lobe epilepsy surgery outcome

Quantitative volumetric brain MRI measurement is important in research applications, but translating it into patient care is challenging. We explore the incorporation of clinical automated quantitative MRI measurements in statistical models predicting outcomes of surgery for temporal lobe epilepsy. Four hundred and thirty-five patients with drug-resistant epilepsy who underwent temporal lobe surgery at Cleveland Clinic, Mayo Clinic and University of Campinas were studied. We obtained volumetric measurements from the pre-operative T1-weighted MRI using NeuroQuant, a Food and Drug Administration approved software package. We created sets of statistical models to predict the probability of complete seizure-freedom or an Engel score of I at the last follow-up. The cohort was randomly split into training and testing sets, with a ratio of 7:3. Model discrimination was assessed using the concordance statistic (C-statistic). We compared four sets of models and selected the one with the highest concordance index. Volumetric differences in pre-surgical MRI located predominantly in the frontocentral and temporal regions were associated with poorer outcomes. The addition of volumetric measurements to the model with clinical variables alone increased the model’s C-statistic from 0.58 to 0.70 (right-sided surgery) and from 0.61 to 0.66 (left-sided surgery) for complete seizure freedom and from 0.62 to 0.67 (right-sided surgery) and from 0.68 to 0.73 (left-sided surgery) for an Engel I outcome score. 57% of patients with extra-temporal abnormalities were seizure-free at last follow-up, compared to 68% of those with no such abnormalities (P-value = 0.02). Adding quantitative MRI data increases the performance of a model developed to predict post-operative seizure outcomes. The distribution of the regions of interest included in the final model supports the notion that focal epilepsies are network disorders and that subtle cortical volume loss outside the surgical site influences seizure outcome.

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.

Temporal lobe epilepsy with or without hippocampal sclerosis: Structural and functional connectivity using advanced MRI techniques

BACKGROUND AND PURPOSE

The aim of this study was to investigate the differences in structural connectivity based on diffusion tensor imaging (DTI) and functional connectivity based on arterial spin labeling (ASL) MRI between temporal lobe epilepsy (TLE) patients with and without hippocampal sclerosis (HS).

METHODS

We enrolled 50 patients with TLE, including 25 patients with HS and 25 patients without HS, who underwent brain MRI, including DTI and ASL. We calculated the network parameters of structural connectivity based on DTI and functional connectivity based on ASL using a graph theoretical analysis. The parameters included global network measures (radius, diameter, characteristic path length, global efficiency, local efficiency, mean clustering coefficient, transitivity, assortative coefficient, and small-worldness index) and a local network measure (betweenness centrality).

RESULTS

The global and local network measures of structural connectivity were not different between TLE patients with and without HS. However, significant differences in functional connectivity existed between the two groups. The radius and diameter of the global network measures in the TLE patients with HS were significantly increased compared with those without HS (4.140 vs. 3.140, p = 0.045; 6.812 vs. 5.132, p = 0.049; respectively). No differences were detected between other global network measures of functional connectivity and local network measure.

CONCLUSIONS

Significant differences in global network measures of functional connectivity based on ASL existed between TLE patients with and without HS. These findings suggest that TLE patients with HS exhibit a more disconnected functional brain network than those without HS.

An Ensemble of Hyperdimensional Classifiers: Hardware-Friendly Short-Latency Seizure Detection With Automatic iEEG Electrode Selection

We propose a new algorithm for detecting epileptic seizures. Our algorithm first extracts three features, namely mean amplitude, line length, and local binary patterns that are fed to an ensemble of classifiers using hyperdimensional (HD) computing. These features are embedded into prototype vectors representing ictal (during seizures) and interictal (between seizures) brain states are constructed. These vectors can be computed at different spatial scales ranging from a single electrode up to many electrodes. This flexibility allows our algorithm to identify the electrodes that discriminate best between ictal and interictal brain states. We assess our algorithm on the SWEC-ETHZ iEEG dataset that includes 99 short-time iEEG seizures recorded with 36 to 100 electrodes from 16 drug-resistant epilepsy patients. Using k-fold cross-validation and all electrodes, our algorithm surpasses state-of-the-art algorithms yielding significantly shorter latency (8.81 s vs. 11.57 s) in seizure onset detection, and higher specificity (97.31% vs. 94.84%) and accuracy (96.85% vs. 95.42%). We can further reduce the latency of our algorithm to 3.74 s by allowing a slightly higher percentage of false alarms (2% specificity loss). Using only the top 10% of the electrodes ranked by our algorithm, we still maintain superior latency, sensitivity, and specificity compared to the other algorithms with all the electrodes. We finally demonstrate the suitability of our algorithm to deployment on low-cost embedded hardware platforms, thanks to its robustness to noise/artifacts affecting the signal, its low computational complexity, and the small memory-footprint on a RISC-V microcontroller.

Recent Advances in Neuroimaging of Epilepsy

Human neuroimaging has had a major impact on the biological understanding of epilepsy and the relationship between pathophysiology, seizure management, and outcomes. This review highlights notable recent advancements in hardware, sequences, methods, analyses, and applications of human neuroimaging techniques utilized to assess epilepsy. These structural, functional, and metabolic assessments include magnetic resonance imaging (MRI), positron emission tomography (PET), and magnetoencephalography (MEG). Advancements that highlight non-invasive neuroimaging techniques used to study the whole brain are emphasized due to the advantages these provide in clinical and research applications. Thus, topics range across presurgical evaluations, understanding of epilepsy as a network disorder, and the interactions between epilepsy and comorbidities. New techniques and approaches are discussed which are expected to emerge into the mainstream within the next decade and impact our understanding of epilepsies. Further, an increasing breadth of investigations includes the interplay between epilepsy, mental health comorbidities, and aberrant brain networks. In the final section of this review, we focus on neuroimaging studies that assess bidirectional relationships between mental health comorbidities and epilepsy as a model for better understanding of the commonalities between both conditions.

Multivariate white matter alterations are associated with epilepsy duration

Previous studies investigating associations between white matter alterations and duration of temporal lobe epilepsy (TLE) have shown differing results, and were typically limited to univariate analyses of tracts in isolation. In this study, we apply a multivariate measure (the Mahalanobis distance), which captures the distinct ways white matter may differ in individual patients, and relate this to epilepsy duration. Diffusion MRI, from a cohort of 94 subjects (28 healthy controls, 33 left-TLE and 33 right-TLE), was used to assess the association between tract fractional anisotropy (FA) and epilepsy duration. Using ten white matter tracts, we analysed associations using the traditional univariate analysis (z-scores) and a complementary multivariate approach (Mahalanobis distance), incorporating multiple white matter tracts into a single unified analysis. For patients with right-TLE, FA was not significantly associated with epilepsy duration for any tract studied in isolation. For patients with left-TLE, the FA of two limbic tracts (ipsilateral fornix, contralateral cingulum gyrus) were significantly negatively associated with epilepsy duration (Bonferonni corrected p < .05). Using a multivariate approach we found significant ipsilateral positive associations with duration in both left, and right-TLE cohorts (left-TLE: Spearman's ρ = 0.487, right-TLE: Spearman's ρ = 0.422). Extrapolating our multivariate results to duration equals zero (i.e., at onset) we found no significant difference between patients and controls. Associations using the multivariate approach were more robust than univariate methods. The multivariate Mahalanobis distance measure provides non-overlapping and more robust results than traditional univariate analyses. Future studies should consider adopting both frameworks into their analysis in order to ascertain a more complete understanding of epilepsy progression, regardless of laterality.

Morphometric and Functional Brain Connectivity Differentiates Chess Masters From Amateur Players

A common task in brain image analysis includes diagnosis of a certain medical condition wherein groups of healthy controls and diseased subjects are analyzed and compared. On the other hand, for two groups of healthy participants with different proficiency in a certain skill, a distinctive analysis of the brain function remains a challenging problem. In this study, we develop new computational tools to explore the functional and anatomical differences that could exist between the brain of healthy individuals identified on the basis of different levels of task experience/proficiency. Toward this end, we look at a dataset of amateur and professional chess players, where we utilize resting-state functional magnetic resonance images to generate functional connectivity (FC) information. In addition, we utilize T1-weighted magnetic resonance imaging to estimate morphometric connectivity (MC) information. We combine functional and anatomical features into a new connectivity matrix, which we term as the functional morphometric similarity connectome (FMSC). Since, both the FC and MC information is susceptible to redundancy, the size of this information is reduced using statistical feature selection. We employ off-the-shelf machine learning classifier, support vector machine, for both single- and multi-modality classifications. From our experiments, we establish that the saliency and ventral attention network of the brain is functionally and anatomically different between two groups of healthy subjects (chess players). We argue that, since chess involves many aspects of higher order cognition such as systematic thinking and spatial reasoning and the identified network is task-positive to cognition tasks requiring a response, our results are valid and supporting the feasibility of the proposed computational pipeline. Moreover, we quantitatively validate an existing neuroscience hypothesis that learning a certain skill could cause a change in the brain (functional connectivity and anatomy) and this can be tested via our novel FMSC algorithm.

Predicting Long-Term Outcomes After Poor-Grade Aneurysmal Subarachnoid Hemorrhage Using Decision Tree Modeling

BACKGROUND

Despite advances in the treatment of poor-grade aneurysmal subarachnoid hemorrhage (aSAH), predicting the long-term outcome of aSAH remains challenging, although essential.

OBJECTIVE

To predict long-term outcomes after poor-grade aSAH using decision tree modeling.

METHODS

This was a retrospective analysis of a prospective multicenter observational registry of patients with poor-grade aSAH with a World Federation of Neurosurgical Societies (WFNS) grade IV or V. Outcome was assessed by the modified Rankin Scale (mRS) at 12 mo, and an unfavorable outcome was defined as an mRS of 4 or 5 or death. Long-term prognostic models were developed using multivariate logistic regression and decision tree algorithms. An additional independent testing dataset was collected for external validation. Overall accuracy, sensitivity, specificity, and area under receiver operating characteristic curves (AUC) were used to assess model performance.

RESULTS

Of the 266 patients, 139 (52.3%) had an unfavorable outcome. Older age, absence of pupillary reactivity, lower Glasgow coma score (GCS), and higher modified Fisher grade were independent predictors of unfavorable outcome. Modified Fisher grade, pupillary reactivity, GCS, and age were used in the decision tree model, which achieved an overall accuracy of 0.833, sensitivity of 0.821, specificity of 0.846, and AUC of 0.88 in the internal test. There was similar predictive performance between the logistic regression and decision tree models. Both models achieved a high overall accuracy of 0.895 in the external test.

CONCLUSION

Decision tree model is a simple tool for predicting long-term outcomes after poor-grade aSAH and may be considered for treatment decision-making.

Altered structural connectome in non-lesional newly diagnosed focal epilepsy: Relation to pharmacoresistance

Despite an expanding literature on brain alterations in patients with longstanding epilepsy, few neuroimaging studies investigate patients with newly diagnosed focal epilepsy (NDfE). Understanding brain network impairments at diagnosis is necessary to elucidate whether or not brain abnormalities are principally due to the chronicity of the disorder and to develop prognostic markers of treatment outcome. Most adults with NDfE do not have MRI-identifiable lesions and the reasons for seizure onset and refractoriness are unknown. We applied structural connectomics to T1-weighted and multi-shell diffusion MRI data with generalized q-sampling image reconstruction using Network Based Statistics (NBS). We scanned 27 patients within an average of 3.7 (SD = 2.9) months of diagnosis and anti-epileptic drug treatment outcomes were collected 24 months after diagnosis. Seven patients were excluded due to lesional NDfE and outcome data was available in 17 patients. Compared to 29 healthy controls, patients with non-lesional NDfE had connectomes with significantly decreased quantitative anisotropy in edges connecting right temporal, frontal and thalamic nodes and increased diffusivity in edges between bilateral temporal, frontal, occipital and parietal nodes. Compared to controls, patients with persistent seizures showed the largest effect size (|d|>=1) for decreased anisotropy in right parietal edges and increased diffusivity in edges between left thalamus and left parietal nodes. Compared to controls, patients who were rendered seizure-free showed the largest effect size for decreased anisotropy in the edge connecting the left thalamus and right temporal nodes and increased diffusivity in edges connecting right frontal nodes. As demonstrated by large effect sizes, connectomes with decreased anisotropy (edge between right frontal and left insular nodes) and increased diffusivity (edge between right thalamus and left parietal nodes) were found in patients with persistent seizures compared to patients who became seizure-free. Patients who had persistent seizures showed larger effect sizes in all network metrics than patients who became seizure-free when compared to each other and compared to controls. Furthermore, patients with focal-to-bilateral tonic-clonic seizures (FBTCS, N = 11) had decreased quantitative anisotropy in a bilateral network involving edges between temporal, parietal and frontal nodes with greater effect sizes than those of patients without FBTCS (N = 9). NBS findings between patients and controls indicated that structural network changes are not necessarily a consequence of longstanding refractory epilepsy and instead are present at the time of diagnosis. Computed effect sizes suggest that there may be structural network MRI-markers of future pharmacoresistance and seizure severity in patients with a new diagnosis of focal epilepsy.

Structural Brain Network Abnormalities and the Probability of Seizure Recurrence After Epilepsy Surgery

OBJECTIVE

We assessed preoperative structural brain networks and clinical characteristics of patients with drug-resistant temporal lobe epilepsy (TLE) to identify correlates of postsurgical seizure recurrences.

METHODS

We examined data from 51 patients with TLE who underwent anterior temporal lobe resection (ATLR) and 29 healthy controls. For each patient, using the preoperative structural, diffusion, and postoperative structural MRI, we generated 2 networks: presurgery network and surgically spared network. Standardizing these networks with respect to controls, we determined the number of abnormal nodes before surgery and expected to be spared by surgery. We incorporated these 2 abnormality measures and 13 commonly acquired clinical data from each patient into a robust machine learning framework to estimate patient-specific chances of seizures persisting after surgery.

RESULTS

Patients with more abnormal nodes had a lower chance of complete seizure freedom at 1 year and, even if seizure-free at 1 year, were more likely to relapse within 5 years. The number of abnormal nodes was greater and their locations more widespread in the surgically spared networks of patients with poor outcome than in patients with good outcome. We achieved an area under the curve of 0.84 ± 0.06 and specificity of 0.89 ± 0.09 in predicting unsuccessful seizure outcomes (International League Against Epilepsy [ILAE] 3-5) as opposed to complete seizure freedom (ILAE 1) at 1 year. Moreover, the model-predicted likelihood of seizure relapse was significantly correlated with the grade of surgical outcome at year 1 and associated with relapses up to 5 years after surgery.

CONCLUSION

Node abnormality offers a personalized, noninvasive marker that can be combined with clinical data to better estimate the chances of seizure freedom at 1 year and subsequent relapse up to 5 years after ATLR.

CLASSIFICATION OF EVIDENCE

This study provides Class II evidence that node abnormality predicts postsurgical seizure recurrence.