Predicting drug-resistant epilepsy - A machine learning approach based on administrative claims data

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.

Comparing the efficacy of anti-seizure medications using matched cohorts on a large insurance claims database

Epilepsy is a severe chronic neurological disease affecting 60 million people worldwide. Primary treatment is with anti-seizure medicines (ASMs), but many patients continue to experience seizures. We used retrospective insurance claims data on 280,587 patients with uncontrolled epilepsy (UE), defined as status epilepticus, need for a rescue medicine, or admission or emergency visit for an epilepsy code. We conducted a computational risk ratio analysis between pairs of ASMs using a causal inference method, in order to match 1034 clinical factors and simulate randomization. Data was extracted from the MarketScan insurance claims Research Database records from 2011 to 2015. The cohort consisted of individuals over 18 years old with a diagnosis of epilepsy who took one of eight ASMs and had more than a year of history prior to the filling of the drug prescription. Seven ASM exposures were analyzed: topiramate, phenytoin, levetiracetam, gabapentin, lamotrigine, valproate, and carbamazepine or oxcarbazepine (treated as the same exposure). We calculated the risk ratio of UE between pairs of ASM after controlling for bias with inverse propensity weighting applied to 1034 factors, such as demographics, confounding illnesses, non-epileptic conditions treated by ASMs, etc. All ASMs exhibited a significant reduction in the prevalence of UE, but three drugs showed pair-wise differences compared to other ASMs. Topiramate consistently was associated with a lower risk of UE, with a mean risk ratio range of 0.68-0.93 (average 0.82, CI: 0.56-1.08). Phenytoin and levetiracetam were consistently associated with a higher risk of UE with mean risk ratio ranges of 1.11 to 1.47 (average 1.13, CI 0.98-1.65) and 1.15 to 1.43 (average 1.2, CI 0.72-1.69), respectively. Large-scale retrospective insurance claims data - combined with causal inference analysis - provides an opportunity to compare the effect of treatments in real-world data in populations 1,000-fold larger than those in typical randomized trials. Our causal analysis identified the clinically unexpected finding of topiramate as being associated with a lower risk of UE; and phenytoin and levetiracetam as associated with a higher risk of UE (compared to other studied drugs, not to baseline). However, we note that our data set for this study only used insurance claims events, which does not comprise actual seizure frequencies, nor a clear picture of side effects. Our results do not advocate for any change in practice but demonstrate that conclusions from large databases may differ from and supplement those of randomized trials and clinical practice and therefore may guide further investigation.

Assessment of potential transthyretin amyloid cardiomyopathy cases in the Brazilian public health system using a machine learning model

OBJECTIVES

To identify and describe the profile of potential transthyretin cardiac amyloidosis (ATTR-CM) cases in the Brazilian public health system (SUS), using a predictive machine learning (ML) model.

METHODS

This was a retrospective descriptive database study that aimed to estimate the frequency of potential ATTR-CM cases in the Brazilian public health system using a supervised ML model, from January 2015 to December 2021. To build the model, a list of ICD-10 codes and procedures potentially related with ATTR-CM was created based on literature review and validated by experts.

RESULTS

From 2015 to 2021, the ML model classified 262 hereditary ATTR-CM (hATTR-CM) and 1,581 wild-type ATTR-CM (wtATTR-CM) potential cases. Overall, the median age of hATTR-CM and wtATTR-CM patients was 66.8 and 59.9 years, respectively. The ICD-10 codes most presented as hATTR-CM and wtATTR-CM were related to heart failure and arrythmias. Regarding the therapeutic itinerary, 13% and 5% of hATTR-CM and wtATTR-CM received treatment with tafamidis meglumine, respectively, while 0% and 29% of hATTR-CM and wtATTR-CM were referred to heart transplant.

CONCLUSION

Our findings may be useful to support the development of health guidelines and policies to improve diagnosis, treatment, and to cover unmet medical needs of patients with ATTR-CM in Brazil.

Early prediction of drug-resistant epilepsy using clinical and EEG features based on convolutional neural network

OBJECTIVE

Machine learning utilization in electroencephalogram (EEG) analysis and epilepsy care is fast evolving. Thus, we aim to develop and validate two one-dimensional convolutional neural network (CNN) algorithms for predicting drug-resistant epilepsy (DRE) in patients with newly-diagnosed epilepsy based on EEG and clinical features.

METHODS

We included a total of 1010 EEG signal epochs and 15 clinical features from 101 patients with epilepsy. Each patient had 10 epochs of EEG signal data, with each signal recorded for 90 s. The ratio of development set and validation set was 80:20, and ten-fold cross validation was performed. First, a CNN algorithm was used to extract EEG features automatically. Then, Two one-dimensional CNNs were crafted.. Accuracy, specificity, precision, sensitivity, F1-score, kappa statistics, mean square error (MSE) and area under the curve (AUC) were calculated to evaluate the classifiers performance.

RESULTS

The clinical-EEG model showed good performance and clinical practical value, with the accuracy, specificity, precision, sensitivity, F1-score, kappa statistics, best MSE and AUC in test set were 0.99, 0.72, 0.82, 0.96, 0.89, 0.83, 32.00, 0.81, respectively, and the accuracy in validation set was 0.84. In the EEG model, the accuracy, specificity, precision, sensitivity, F1-score, kappa statistics, best MSE and AUC in test set were 0.99, 0.59, 0.82, 0.90, 0.86, 0.72, 181.76, 0.76, respectively, and the accuracy in validation set was 0.81.

CONCLUSION

We constructed a clinical-EEG model showed good potential for predicting DRE in patients with newly-diagnosed epilepsy, which could help identify patients at high risk of developing DRE at earlier stages.

Predicting Antiseizure Medication Treatment in Children with Rare Tuberous Sclerosis Complex-Related Epilepsy Using Deep Learning

BACKGROUND AND PURPOSE

Tuberous sclerosis complex disease is a rare, multisystem genetic disease, but appropriate drug treatment allows many pediatric patients to have positive outcomes. The purpose of this study was to predict the effectiveness of antiseizure medication treatment in children with tuberous sclerosis complex-related epilepsy.

MATERIALS AND METHODS

We conducted a retrospective study involving 300 children with tuberous sclerosis complex-related epilepsy. The study included the analysis of clinical data and T2WI and FLAIR images. The clinical data consisted of sex, age of onset, age at imaging, infantile spasms, and antiseizure medication numbers. To forecast antiseizure medication treatment, we developed a multitechnique deep learning method called WAE-Net. This method used multicontrast MR imaging and clinical data. The T2WI and FLAIR images were combined as FLAIR3 to enhance the contrast between tuberous sclerosis complex lesions and normal brain tissues. We trained a clinical data-based model using a fully connected network with the above-mentioned variables. After that, a weighted-average ensemble network built from the ResNet3D architecture was created as the final model.

RESULTS

The experiments had shown that age of onset, age at imaging, infantile spasms, and antiseizure medication numbers were significantly different between the 2 drug-treatment outcomes (P < .05). The hybrid technique of FLAIR3 could accurately localize tuberous sclerosis complex lesions, and the proposed method achieved the best performance (area under the curve = 0.908 and accuracy of 0.847) in the testing cohort among the compared methods.

CONCLUSIONS

The proposed method could predict antiseizure medication treatment of children with rare tuberous sclerosis complex-related epilepsy and could be a strong baseline for future studies.

Drug resistance in epilepsy

Drug resistance is estimated to affect about a third of individuals with epilepsy, but its prevalence differs in relation to the epilepsy syndrome, the cause of epilepsy, and other factors such as age of seizure onset and presence of associated neurological deficits. Although drug-resistant epilepsy is not synonymous with unresponsiveness to any drug treatment, the probability of achieving seizure freedom on a newly tried medication decreases with increasing number of previously failed treatments. After two appropriately used antiseizure medications have failed to control seizures, individuals should be referred whenever possible to a comprehensive epilepsy centre for diagnostic re-evaluation and targeted management. The feasibility of epilepsy surgery and other treatments, including those targeting the cause of epilepsy, should be considered early after diagnosis. Substantial evidence indicates that a delay in identifying an effective treatment can adversely affect ultimate outcome and carry an increased risk of cognitive disability, other comorbidities, and premature mortality. Research on mechanisms of drug resistance and novel therapeutics is progressing rapidly, and potentially improved treatments, including those targeting disease modification, are on the horizon.

Variation in prognosis and treatment outcome in juvenile myoclonic epilepsy: a Biology of Juvenile Myoclonic Epilepsy Consortium proposal for a practical definition and stratified medicine classifications

Reliable definitions, classifications and prognostic models are the cornerstones of stratified medicine, but none of the current classifications systems in epilepsy address prognostic or outcome issues. Although heterogeneity is widely acknowledged within epilepsy syndromes, the significance of variation in electroclinical features, comorbidities and treatment response, as they relate to diagnostic and prognostic purposes, has not been explored. In this paper, we aim to provide an evidence-based definition of juvenile myoclonic epilepsy showing that with a predefined and limited set of mandatory features, variation in juvenile myoclonic epilepsy phenotype can be exploited for prognostic purposes. Our study is based on clinical data collected by the Biology of Juvenile Myoclonic Epilepsy Consortium augmented by literature data. We review prognosis research on mortality and seizure remission, predictors of antiseizure medication resistance and selected adverse drug events to valproate, levetiracetam and lamotrigine. Based on our analysis, a simplified set of diagnostic criteria for juvenile myoclonic epilepsy includes the following: (i) myoclonic jerks as mandatory seizure type; (ii) a circadian timing for myoclonia not mandatory for the diagnosis of juvenile myoclonic epilepsy; (iii) age of onset ranging from 6 to 40 years; (iv) generalized EEG abnormalities; and (v) intelligence conforming to population distribution. We find sufficient evidence to propose a predictive model of antiseizure medication resistance that emphasises (i) absence seizures as the strongest stratifying factor with regard to antiseizure medication resistance or seizure freedom for both sexes and (ii) sex as a major stratifying factor, revealing elevated odds of antiseizure medication resistance that correlates to self-report of catamenial and stress-related factors including sleep deprivation. In women, there are reduced odds of antiseizure medication resistance associated with EEG-measured or self-reported photosensitivity. In conclusion, by applying a simplified set of criteria to define phenotypic variations of juvenile myoclonic epilepsy, our paper proposes an evidence-based definition and prognostic stratification of juvenile myoclonic epilepsy. Further studies in existing data sets of individual patient data would be helpful to replicate our findings, and prospective studies in inception cohorts will contribute to validate them in real-world practice for juvenile myoclonic epilepsy management.

Individualized Prediction of Drug Resistance in People with Post-Stroke Epilepsy: A Retrospective Study

BACKGROUND

The study aimed to develop a model and build a nomogram to predict the probability of drug resistance in people with post-stroke epilepsy (PSE).

METHODS

Subjects with epilepsy secondary to ischemic stroke or spontaneous intracerebral hemorrhage were included. The study outcome was the occurrence of drug-resistant epilepsy defined according to International League Against Epilepsy criteria.

RESULTS

One hundred and sixty-four subjects with PSE were included and 32 (19.5%) were found to be drug-resistant. Five variables were identified as independent predictors of drug resistance and were included in the nomogram: age at stroke onset (odds ratio (OR): 0.941, 95% confidence interval (CI) 0.907-0.977), intracerebral hemorrhage (OR: 6.292, 95% CI 1.957-20.233), severe stroke (OR: 4.727, 95% CI 1.573-14.203), latency of PSE (>12 months, reference; 7-12 months, OR: 4.509, 95% CI 1.335-15.228; 0-6 months, OR: 99.099, 95% CI 14.873-660.272), and status epilepticus at epilepsy onset (OR: 14.127, 95% CI 2.540-78.564). The area under the receiver operating characteristic curve of the nomogram was 0.893 (95% CI: 0.832-0.956).

CONCLUSIONS

Great variability exists in the risk of drug resistance in people with PSE. A nomogram based on a set of readily available clinical variables may represent a practical tool for an individualized prediction of drug-resistant PSE.

eTSC-Net: A Parameter-efficient Convolutional Neural Network for Drug Treatment Outcome Studies of Pediatric Epilepsy.. [DOI: 10.21203/rs.3.rs-2024294/v1] [Citation(s) in RCA: 0] [Impact Index Per Article: 0]

Background: Ability to predict the outcomes of pharmacological treatment of epilepsy in pediatric patients with tuberous sclerosis complex (TSC) can confer a distinct leverage and guide therapeutic decision-making. Multi-contrast magnetic resonance imaging (MRI) is routinely used for diagnosis of TSC by pediatricians. We propose a parameter-efficient convolutional neural network with multi-contrast images to predict the drug treatment outcomes of pediatric epilepsy in TSC. Methods: Image-based models were generated using the EfficientNet3D-B0 network architecture. A weighted average ensemble network with multi-contrast images was created as the final model. The proposed neural network is named as Efficient Tuberous sclerosis complex-Net (eTSC-Net).We compared our methods with a Residual Network 3D（ResNet3D) model. We trained a 3D-ResNet on our T2FLAIR data. Binary classification models were trained to distinguish non-controlled group patients from controlled group patients on T2W and T2FLAIR images. We trained all the models using an Nvidia RTX A6000 Graphical Processing Unit (GPU) card. Area under curve(AUC), accuracy (ACC), sensitivity (SEN), and specificity (SPE) were calculated to assess the classification performance for each model in each cohort. The differences between subgroups were assessed using independent samples t test and pvalues < 0.05 were considered indicative of statistical significance. Results: The proposed neural network (eTSC-Net) achieved the best performance with an AUC value of 0.833 and 90.0% accuracy in the testing cohort, which was better than other models. Conclusions: The results demonstrated the ability of the proposed method for predicting drug treatment outcomes in pediatric TSC-related epilepsy. eTSC-Net can serve as a useful computer-aided diagnostic tool to help clinical radiologists formulate more targeted treatment.

Machine learning and statistic analysis to predict drug treatment outcome in pediatric epilepsy patients with tuberous sclerosis complex

OBJECTIVES

We aimed to investigate the association between multi-modality features and epilepsy drug treatment outcomes and propose a machine learning model to predict epilepsy drug treatment outcomes with multi-modality features.

METHODS

This retrospective study consecutively enrolled 103 epilepsy children with rare TSC. Multi-modality data were used to characterize risk factors for epilepsy drug treatment outcome of TSC, including clinical data, TSC1, and TSC2 genes test results, magnetic resonance imaging (MRI), computerized tomography (CT), and electroencephalogram (EEG). Three common feature selection methods and six common machine learning models were used to find the best combination of feature selection and machine learning model for epilepsy drug treatment outcomes prediction with multi-modality features for TSC clinical application.

RESULTS

The analysis of variance based on selected 35 features combined with multilayer perceptron (MLP) model achieved the best area-under-curve score (AUC) of 0.812 (±0.005). Infantile spasms, EEG discharge type, epileptiform discharge in the right frontal area of EEG, drug-resistant epilepsy, gene mutation type, and type II lesions were positively correlated with drug treatment outcome. Age of onset and age of visiting doctors were negatively correlated with drug treatment outcome (p < 0.05). Our machine learning results found that among MRI features, lesion type is the most important in the outcome prediction, followed by location and quantity.

CONCLUSION

We developed and validated an effective prediction model for epilepsy drug treatment outcomes of TSC. Our results suggested that multi-modality features analysis and MLP-based machine learning can predict epilepsy drug treatment outcomes of TSC.

A machine-learning guided method for predicting add-on and switch in secondary data sources: A case study on anti-seizure medications in Danish registries

Purpose: There is a lack of available evidence regarding the treatment pattern of switches and add-ons for individuals aged 65 years or older with epilepsy during the first years from the time they received their first anti-seizure medication because of the lack of valid methods. Therefore, this study aimed to develop an algorithm for identifying switches and add-ons using secondary data sources for anti-seizure medication users. Methods: Danish nationwide databases were used as data sources. Residents in Denmark between 1996 and 2018 who were diagnosed with epilepsy and redeemed their first prescription for anti-seizure medication after epilepsy diagnosis were followed up for 730 days until the end of the follow-up period, death, or emigration to assess switches and add-ons occurred during the follow-up period. The study outcomes were the overall accuracy of the classification of switch or add-on of the newly developed algorithm. Results: In total, 15870 individuals were included in the study population with a median age of 72.9 years, of whom 52.0% were male and 48.0% were female. A total of 988 of the 15879 patients from the study population were present during the 730-day follow-up period, and 988 individuals (6.2%) underwent a total of 1485 medication events with co-exposure to two or more anti-seizure medications. The newly developed algorithmic method correctly identified 9 out of 10 add-ons (overall accuracy 92%) and 9 out of 10 switches (overall accuracy 88%). Conclusion: The majority of switches and add-ons occurred early during the first 2 years of disease and according to clinical recommendations. The newly developed algorithm correctly identified 9 out of 10 switches/add-ons.

Application of Machine Learning Methods for Epilepsy Risk Ranking in Patients with Hematopoietic Malignancies Using

Machine learning methods to predict the risk of epilepsy, including vascular epilepsy, in oncohematological patients are currently considered promising. These methods are used in research to predict pharmacoresistant epilepsy and surgical treatment outcomes in order to determine the epileptogenic zone and functional neural systems in patients with epilepsy, as well as to develop new approaches to classification and perform other tasks. This paper presents the results of applying machine learning to analyzing data and developing diagnostic models of epilepsy in oncohematological and cardiovascular patients. This study contributes to solving the problem of often unjustified diagnosis of primary epilepsy in patients with oncohematological or cardiovascular pathology, prescribing antiseizure drugs to patients with single seizure syndromes without finding a disease associated with these cases. We analyzed the hospital database of the V.A. Almazov Scientific Research Center of the Ministry of Health of Russia. The study included 66,723 treatment episodes of patients with vascular diseases (I10–I15, I61–I69, I20–I25) and 16,383 episodes with malignant neoplasms of lymphoid, hematopoietic, and related tissues (C81–C96 according to ICD-10) for the period from 2010 to 2020. Data analysis and model calculations indicate that the best result was shown by gradient boosting with mean accuracy cross-validation score = 0.96. f1-score = 98, weighted avg precision = 93, recall = 96, f1-score = 94. The highest correlation coefficient for G40 and different clinical conditions was achieved with fibrillation, hypertension, stenosis or occlusion of the precerebral arteries (0.16), cerebral sinus thrombosis (0.089), arterial hypertension (0.17), age (0.03), non-traumatic intracranial hemorrhage (0.07), atrial fibrillation (0.05), delta absolute neutrophil count (0.05), platelet count at discharge (0.04), transfusion volume for stem cell transplantation (0.023). From the clinical point of view, the identified differences in the importance of predictors in a broader patient model are consistent with a practical algorithm for organic brain damage. Atrial fibrillation is one of the leading factors in the development of both ischemic and hemorrhagic strokes. At the same time, brain infarction can be accompanied both by the development of epileptic seizures in the acute period and by unprovoked epileptic seizures and development of epilepsy in the early recovery and in a longer period. In addition, a microembolism of the left heart chambers can lead to multiple microfocal lesions of the brain, which is one of the pathogenetic aspects of epilepsy in elderly patients. The presence of precordial fibrillation requires anticoagulant therapy, the use of which increases the risk of both spontaneous and traumatic intracranial hemorrhage.

Differences between men and women in response to antiseizure medication use and the likelihood of developing treatment resistant epilepsy

OBJECTIVE

The prevalence of epilepsy is slightly higher in women than in men and sensitivity to seizure stimuli differs between sexes. Some evidence suggests sex differences in response to antiseizure medications exist mainly due to inconsistent pharmacokinetic differences; however, there is a lack of real-world evidence examining differences in response to antiseizure medications between men and women.

METHODS

This was a retrospective population-based cohort study in five large US healthcare databases. The population included adult patients with epilepsy, newly exposed to levetiracetam, and naive to antiseizure medication. The first exposure to levetiracetam was the index date. The requirement that all patients received the same medication was done to avoid potential confounding due to differences in index treatment. The outcome was the development of treatment resistant epilepsy (TRE), defined as having at least three distinct antiseizure medications in 1 year. The proportion of patients who developed TRE within 1 year following the index date was calculated. To compare the risk of developing TRE between sexes, relative risks (RR) and 95% confidence intervals (CI) were calculated, and estimates were pooled using meta-analytic techniques stratified by gender and age.

RESULTS

A total of 147 334 subjects were included in the databases, 50.8% were women, and 4.27% developed TRE. The comorbid profile differed greatly between men and women; however, the types of epilepsy syndromes observed during baseline were similar between the two groups. Across all databases, women were more likely to develop TRE than men (pooled RR 1.27, 95% CI 1.17-1.38). Results remained similar when stratified by age.

SIGNIFICANCE

This study assessed sex differences in response to antiseizure medications using the development of TRE as a proxy for effectiveness. Women newly exposed to levetiracetam were 27% more likely to develop TRE than men, independent of age.

Analysis of Patients with Focal Epilepsy and Drug-Resistant Epilepsy in Italy: Evaluation of Their Characteristics, Therapeutic Pathway and the Consumption of Healthcare Resources

Purpose

A retrospective analysis was conducted to estimate the number of patients with focal epilepsy and drug-resistant epilepsy (DRE) and their characteristics, the therapeutic patterns, the consumption of health resources in a real-world Italian setting.

Patients and Methods

A retrospective study was carried out on the administrative databases of a sample of Italian Health Departments, covering approximately 8.7 million health-assisted individuals. All adult patients with at least one hospitalization for focal epilepsy and an electroencephalogram (between 01/2010 and 12/2019), and at least one prescription of antiseizure medication (ASM) (between 01/2011 and 12/2018) were included in the study. Patients with at least two treatment failures and treated with a subsequent ASM were considered DRE.

Results

Overall, 1897 patients with focal epilepsy (mean age 56 years, 47% male) were identified, of which 485 (25.6%) with DRE (mean age 53 years, 43% male). Among patients with focal epilepsy and DRE, respectively, 48% and 54% had essential hypertension, 23.4% and 26.6% had cardiovascular disease, and 46.3% and 62.1% had peptic ulcer/prescription of gastric secretion inhibitors. During follow-up, patients with focal epilepsy maintained first-line treatment for 53.9 months; among these, 52% passed to the second-line, and 485 (25.6% of the total) began third-line treatment. In patients with focal epilepsy, the mean cost was € 4448 (of which € 1410 were epilepsy-related), and in DRE patients total expenditures averages € 5825 (of which € 2165 were epilepsy-related). In both patients with focal epilepsy and DRE, hospitalizations represented the most impacting item of expenditure.

Conclusion

The present analysis conducted in a setting of Italian clinical practice has shown that 25% of patients with focal epilepsy were resistant to antiepileptic treatments. Furthermore, these results showed that health-care costs for the management of epileptic patients were mainly accountable for the costs related to the disease-management and to hospitalizations.

Predicting the Mortality of ICU Patients by Topic Model with Machine-Learning Techniques

Predicting clinical patients’ vital signs is a leading critical issue in intensive care units (ICUs) related studies. Early prediction of the mortality of ICU patients can reduce the overall mortality and cost of complication treatment. Some studies have predicted mortality based on electronic health record (EHR) data by using machine learning models. However, the semi-structured data (i.e., patients’ diagnosis data and inspection reports) is rarely used in these models. This study utilized data from the Medical Information Mart for Intensive Care III. We used a Latent Dirichlet Allocation (LDA) model to classify text in the semi-structured data of some particular topics and established and compared the classification and regression trees (CART), logistic regression (LR), multivariate adaptive regression splines (MARS), random forest (RF), and gradient boosting (GB). A total of 46,520 ICU Patients were included, with 11.5% mortality in the Medical Information Mart for Intensive Care III group. Our results revealed that the semi-structured data (diagnosis data and inspection reports) of ICU patients contain useful information that can assist clinical doctors in making critical clinical decisions. In addition, in our comparison of five machine learning models (CART, LR, MARS, RF, and GB), the GB model showed the best performance with the highest area under the receiver operating characteristic curve (AUROC) (0.9280), specificity (93.16%), and sensitivity (83.25%). The RF, LR, and MARS models showed better performance (AUROC are 0.9096, 0.8987, and 0.8935, respectively) than the CART (0.8511). The GB model showed better performance than other machine learning models (CART, LR, MARS, and RF) in predicting the mortality of patients in the intensive care unit. The analysis results could be used to develop a clinically useful decision support system.

Current Principles in the Management of Drug-Resistant Epilepsy

Drug-resistant epilepsy is associated with poor health outcomes and increased economic burden. In the last three decades, various new antiseizure medications have been developed, but the proportion of people with drug-resistant epilepsy remains relatively unchanged. Developing strategies to address drug-resistant epilepsy is essential. Here, we define drug-resistant epilepsy and emphasize its relationship to the conceptualization of epilepsy as a symptom complex, delineate clinical risk factors, and characterize mechanisms based on current knowledge. We address the importance of ruling out pseudoresistance and consider the impact of nonadherence on determining whether an individual has drug-resistant epilepsy. We then review the principles of epilepsy drug therapy and briefly touch upon newly approved and experimental antiseizure medications.

The path from big data analytics capabilities to value in hospitals: a scoping review

BACKGROUND

As the uptake of health information technologies increased, most healthcare organizations have become producers of big data. A growing number of hospitals are investing in the development of big data analytics (BDA) capabilities. If the promises associated with these capabilities are high, how hospitals create value from it remains unclear. The present study undertakes a scoping review of existing research on BDA use in hospitals to describe the path from BDA capabilities (BDAC) to value and its associated challenges.

METHODS

This scoping review was conducted following Arksey and O'Malley's 5 stages framework. A systematic search strategy was adopted to identify relevant articles in Scopus and Web of Science. Data charting and extraction were performed following an analytical framework that builds on the resource-based view of the firm to describe the path from BDA capabilities to value in hospitals.

RESULTS

Of 1,478 articles identified, 94 were included. Most of them are experimental research (n=69) published in medical (n=66) or computer science journals (n=28). The main value targets associated with the use of BDA are improving the quality of decision-making (n=56) and driving innovation (n=52) which apply mainly to care (n=67) and administrative (n=48) activities. To reach these targets, hospitals need to adequately combine BDA capabilities and value creation mechanisms (VCM) to enable knowledge generation and drive its assimilation. Benefits are endpoints of the value creation process. They are expected in all articles but realized in a few instances only (n=19).

CONCLUSIONS

This review confirms the value creation potential of BDA solutions in hospitals. It also shows the organizational challenges that prevent hospitals from generating actual benefits from BDAC-building efforts. The configuring of strategies, technologies and organizational capabilities underlying the development of value-creating BDA solutions should become a priority area for research, with focus on the mechanisms that can drive the alignment of BDA and organizational strategies, and the development of organizational capabilities to support knowledge generation and assimilation.

Development and validation of a nomogram for the early prediction of drug resistance in children with epilepsy

BACKGROUND AND PURPOSE

This study aimed to effectively identify children with drug-resistant epilepsy (DRE) in the early stage of epilepsy, and take personalized interventions, to improve patients' prognosis, reduce serious comorbidity, and save social resources. Herein, we developed and validated a nomogram prediction model for children with DRE.

METHODS

The training set was patients with epilepsy who visited the Children's Hospital of Soochow University (Suzhou Industrial Park, Jiangsu Province, China) between January 2015 and December 2017. The independent risk factors for DRE were screened by univariate and multivariate logistic regression analyses using SPSS21 software. The nomogram was designed according to the regression coefficient. The nomogram was validated in the training and validation sets. Internal validation was conducted using bootstrapping analyses. We also externally validated this instrument in patients with epilepsy from the Children's Hospital of Soochow University (Gusu District, Jiangsu Province, China) and Yancheng Maternal and Child Health Hospital between January 2018 and December 2018. The nomogram's performance was assessed by concordance (C-index), calibration curves, as well as GiViTI calibration belts.

RESULTS

Multivariate logistic regression analysis of 679 children with epilepsy from the Children's Hospital of Soochow University (Suzhou Industrial Park, Jiangsu Province, China) showed that onset age<1, status epilepticus (SE), focal seizure, > 20 pre-treatment seizures, clear etiology (caused by genetic, structural, metabolic, or infectious), development and epileptic encephalopathy (DEE), and neurological abnormalities were all independent risk factors for DRE. The AUC of 0.92 for the training set compared to that of 0.91 for the validation set suggested a good discrimination ability of the prediction model. The C-index was 0.92 and 0.91 in the training and validation sets. Additionally, both good calibration curves and GiViTI calibration belts (P-value: 0.849 and 0.291, respectively) demonstrated that the predicted risks had strong consistency with the observed outcomes, suggesting that the prediction model in both groups was perfectly calibrated.

CONCLUSION

A nomogram prediction model for DRE was developed, with good discrimination and calibration in the training set and the validation set. Furthermore, the model demonstrated great accuracy, consistency, and prediction ability. Therefore, the nomogram prediction model can aid in the timely identification of DRE in children.

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.

Recent trends in artificial intelligence-driven identification and development of anti-neurodegenerative therapeutic agents

Neurological disorders affect various aspects of life. Finding drugs for the central nervous system is a very challenging and complex task due to the involvement of the blood-brain barrier, P-glycoprotein, and the drug's high attrition rates. The availability of big data present in online databases and resources has enabled the emergence of artificial intelligence techniques including machine learning to analyze, process the data, and predict the unknown data with high efficiency. The use of these modern techniques has revolutionized the whole drug development paradigm, with an unprecedented acceleration in the central nervous system drug discovery programs. Also, the new deep learning architectures proposed in many recent works have given a better understanding of how artificial intelligence can tackle big complex problems that arose due to central nervous system disorders. Therefore, the present review provides comprehensive and up-to-date information on machine learning/artificial intelligence-triggered effort in the brain care domain. In addition, a brief overview is presented on machine learning algorithms and their uses in structure-based drug design, ligand-based drug design, ADMET prediction, de novo drug design, and drug repurposing. Lastly, we conclude by discussing the major challenges and limitations posed and how they can be tackled in the future by using these modern machine learning/artificial intelligence approaches.

Definitions of Drug-Resistant Epilepsy for Administrative Claims Data Research

OBJECTIVE

To assess accuracy of definitions of drug-resistant epilepsy applied to administrative claims data.

METHODS

We randomly sampled 450 patients from a tertiary health system with >1 epilepsy/convulsion encounter and >2 distinct antiseizure medications (ASMs) from 2014-2020 and >2 years of electronic medical records (EMR) data. We established a drug-resistant epilepsy diagnosis at a specific visit by reviewing EMR data and employing a rubric based in the 2010 International League Against Epilepsy definition. We performed logistic regressions to assess clinically-relevant predictors of drug-resistant epilepsy and to inform claims-based definitions.

RESULTS

Of 450 patients reviewed, 150 were excluded for insufficient EMR data. Of the 300 patients included, 98 (33%) met criteria for current drug-resistant epilepsy. The strongest predictors of current drug-resistant epilepsy were drug-resistant epilepsy diagnosis code (OR 16.9, 95% CI 8.8-32.2), >2 ASMs in the prior two years (OR 13.0, 95% CI 5.1-33.3), >3 non-gabapentinoid ASMs (OR 10.3, 95% CI 5.4-19.6), neurosurgery visit (OR 45.2, 95% CI 5.9-344.3), and epilepsy surgery (OR 30.7, 95% CI 7.1-133.3). We created claims-based drug-resistant epilepsy definitions to: 1) maximize overall predictiveness (drug-resistant epilepsy diagnosis; sensitivity 0.86, specificity 0.74, area under the receiver operating characteristics curve [AUROC] 0.80), 2) maximize sensitivity (drug-resistant epilepsy diagnosis or >3 ASMs; sensitivity 0.98, specificity 0.47, AUROC 0.72), and 3) maximize specificity (drug-resistant epilepsy diagnosis and >3 non-gabapentinoid ASMs; sensitivity 0.42, specificity 0.98, AUROC 0.70).

CONCLUSIONS

Our findings provide validation for several claims-based definitions of drug-resistant epilepsy that can be applied to a variety of research questions.

Dementia risks identified by vocal features via telephone conversations: A novel machine learning prediction model

Due to difficulty in early diagnosis of Alzheimer's disease (AD) related to cost and differentiated capability, it is necessary to identify low-cost, accessible, and reliable tools for identifying AD risk in the preclinical stage. We hypothesized that cognitive ability, as expressed in the vocal features in daily conversation, is associated with AD progression. Thus, we have developed a novel machine learning prediction model to identify AD risk by using the rich voice data collected from daily conversations, and evaluated its predictive performance in comparison with a classification method based on the Japanese version of the Telephone Interview for Cognitive Status (TICS-J). We used 1,465 audio data files from 99 Healthy controls (HC) and 151 audio data files recorded from 24 AD patients derived from a dementia prevention program conducted by Hachioji City, Tokyo, between March and May 2020. After extracting vocal features from each audio file, we developed machine-learning models based on extreme gradient boosting (XGBoost), random forest (RF), and logistic regression (LR), using each audio file as one observation. We evaluated the predictive performance of the developed models by describing the receiver operating characteristic (ROC) curve, calculating the areas under the curve (AUCs), sensitivity, and specificity. Further, we conducted classifications by considering each participant as one observation, computing the average of their audio files' predictive value, and making comparisons with the predictive performance of the TICS-J based questionnaire. Of 1,616 audio files in total, 1,308 (81.0%) were randomly allocated to the training data and 308 (19.1%) to the validation data. For audio file-based prediction, the AUCs for XGboost, RF, and LR were 0.863 (95% confidence interval [CI]: 0.794-0.931), 0.882 (95% CI: 0.840-0.924), and 0.893 (95%CI: 0.832-0.954), respectively. For participant-based prediction, the AUC for XGboost, RF, LR, and TICS-J were 1.000 (95%CI: 1.000-1.000), 1.000 (95%CI: 1.000-1.000), 0.972 (95%CI: 0.918-1.000) and 0.917 (95%CI: 0.918-1.000), respectively. There was difference in predictive accuracy of XGBoost and TICS-J with almost approached significance (p = 0.065). Our novel prediction model using the vocal features of daily conversations demonstrated the potential to be useful for the AD risk assessment.

An Explainable Multimodal Neural Network Architecture for Predicting Epilepsy Comorbidities Based on Administrative Claims Data

Epilepsy is a complex brain disorder characterized by repetitive seizure events. Epilepsy patients often suffer from various and severe physical and psychological comorbidities (e.g., anxiety, migraine, and stroke). While general comorbidity prevalences and incidences can be estimated from epidemiological data, such an approach does not take into account that actual patient-specific risks can depend on various individual factors, including medication. This motivates to develop a machine learning approach for predicting risks of future comorbidities for individual epilepsy patients. In this work, we use inpatient and outpatient administrative health claims data of around 19,500 U.S. epilepsy patients. We suggest a dedicated multimodal neural network architecture (Deep personalized LOngitudinal convolutional RIsk model-DeepLORI) to predict the time-dependent risk of six common comorbidities of epilepsy patients. We demonstrate superior performance of DeepLORI in a comparison with several existing methods. Moreover, we show that DeepLORI-based predictions can be interpreted on the level of individual patients. Using a game theoretic approach, we identify relevant features in DeepLORI models and demonstrate that model predictions are explainable in light of existing knowledge about the disease. Finally, we validate the model on independent data from around 97,000 patients, showing good generalization and stable prediction performance over time.

A machine learning model for identifying patients at risk for wild-type transthyretin amyloid cardiomyopathy

Transthyretin amyloid cardiomyopathy, an often unrecognized cause of heart failure, is now treatable with a transthyretin stabilizer. It is therefore important to identify at-risk patients who can undergo targeted testing for earlier diagnosis and treatment, prior to the development of irreversible heart failure. Here we show that a random forest machine learning model can identify potential wild-type transthyretin amyloid cardiomyopathy using medical claims data. We derive a machine learning model in 1071 cases and 1071 non-amyloid heart failure controls and validate the model in three nationally representative cohorts (9412 cases, 9412 matched controls), and a large, single-center electronic health record-based cohort (261 cases, 39393 controls). We show that the machine learning model performs well in identifying patients with cardiac amyloidosis in the derivation cohort and all four validation cohorts, thereby providing a systematic framework to increase the suspicion of transthyretin cardiac amyloidosis in patients with heart failure.

Incidence and Prevalence of Drug-Resistant Epilepsy: A Systematic Review and Meta-analysis

OBJECTIVE

To evaluate the incidence and prevalence of drug-resistant epilepsy (DRE) as well as its predictors and correlates, we conducted a systematic review and meta-analysis of observational studies.

METHODS

Our protocol was registered with PROSPERO, and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses and Meta-analysis of Observational Studies in Epidemiology reporting standards were followed. We searched MEDLINE, Embase, and Web of Science. We used a double arcsine transformation and random-effects models to perform our meta-analyses. We performed random-effects meta-regressions using study-level data.

RESULTS

Our search strategy identified 10,794 abstracts. Of these, 103 articles met our eligibility criteria. There was high interstudy heterogeneity and risk of bias. The cumulative incidence of DRE was 25.0% (95% confidence interval [CI]: 16.8-34.3) in child studies but 14.6% (95% CI: 8.8-21.6) in adult/mixed age studies. The prevalence of DRE was 13.7% (95% CI: 9.2-19.0) in population/community-based populations but 36.3% (95% CI: 30.4-42.4) in clinic-based cohorts. Meta-regression confirmed that the prevalence of DRE was higher in clinic-based populations and in focal epilepsy. Multiple predictors and correlates of DRE were identified. The most reported of these were having a neurologic deficit, an abnormal EEG, and symptomatic epilepsy. The most reported genetic predictors of DRE were polymorphisms of the ABCB1 gene.

CONCLUSIONS

Our observations provide a basis for estimating the incidence and prevalence of DRE, which vary between populations. We identified numerous putative DRE predictors and correlates. These findings are important to plan epilepsy services, including epilepsy surgery, a crucial treatment option for people with disabling seizures and DRE.

Machine Learning for Predicting Epileptic Seizures Using EEG Signals: A Review

With the advancement in artificial intelligence (AI) and machine learning (ML) techniques, researchers are striving towards employing these techniques for advancing clinical practice. One of the key objectives in healthcare is the early detection and prediction of disease to timely provide preventive interventions. This is especially the case for epilepsy, which is characterized by recurrent and unpredictable seizures. Patients can be relieved from the adverse consequences of epileptic seizures if it could somehow be predicted in advance. Despite decades of research, seizure prediction remains an unsolved problem. This is likely to remain at least partly because of the inadequate amount of data to resolve the problem. There have been exciting new developments in ML-based algorithms that have the potential to deliver a paradigm shift in the early and accurate prediction of epileptic seizures. Here we provide a comprehensive review of state-of-the-art ML techniques in early prediction of seizures using EEG signals. We will identify the gaps, challenges, and pitfalls in the current research and recommend future directions.

Artificial Intelligence in Pharmacoepidemiology: A Systematic Review. Part 2-Comparison of the Performance of Artificial Intelligence and Traditional Pharmacoepidemiological Techniques

Aim: To summarize the evidence on the performance of artificial intelligence vs. traditional pharmacoepidemiological techniques.

Methods: Ovid MEDLINE (01/1950 to 05/2019) was searched to identify observational studies, meta-analyses, and clinical trials using artificial intelligence techniques having a drug as the exposure or the outcome of the study. Only studies with an available full text in the English language were evaluated.

Results: In all, 72 original articles and five reviews were identified via Ovid MEDLINE of which 19 (26.4%) compared the performance of artificial intelligence techniques with traditional pharmacoepidemiological methods. In total, 44 comparisons have been performed in articles that aimed at 1) predicting the needed dosage given the patient’s characteristics (31.8%), 2) predicting the clinical response following a pharmacological treatment (29.5%), 3) predicting the occurrence/severity of adverse drug reactions (20.5%), 4) predicting the propensity score (9.1%), 5) identifying subpopulation more at risk of drug inefficacy (4.5%), 6) predicting drug consumption (2.3%), and 7) predicting drug-induced lengths of stay in hospital (2.3%). In 22 out of 44 (50.0%) comparisons, artificial intelligence performed better than traditional pharmacoepidemiological techniques. Random forest (seven out of 11 comparisons; 63.6%) and artificial neural network (six out of 10 comparisons; 60.0%) were the techniques that in most of the comparisons outperformed traditional pharmacoepidemiological methods.

Conclusion: Only a small fraction of articles compared the performance of artificial intelligence techniques with traditional pharmacoepidemiological methods and not all artificial intelligence techniques have been compared in a Pharmacoepidemiological setting. However, in 50% of comparisons, artificial intelligence performed better than pharmacoepidemiological techniques.

Application of Artificial Intelligence in Pharmaceutical and Biomedical Studies

BACKGROUND

Artificial intelligence (AI) is the way to model human intelligence to accomplish certain tasks without much intervention of human beings. The term AI was first used in 1956 with The Logic Theorist program, which was designed to simulate problem-solving ability of human beings. There have been a significant amount of research works using AI in order to determine the advantages and disadvantages of its applicabication and, future perspectives that impact different areas of society. Even the remarkable impact of AI can be transferred to the field of healthcare with its use in pharmaceutical and biomedical studies crucial for the socioeconomic development of the population in general within different studies, we can highlight those that have been conducted with the objective of treating diseases, such as cancer, neurodegenerative diseases, among others. In parallel, the long process of drug development also requires the application of AI to accelerate research in medical care.

METHODS

This review is based on research material obtained from PubMed up to Jan 2020. The search terms include "artificial intelligence", "machine learning" in the context of research on pharmaceutical and biomedical applications.

RESULTS

This study aimed to highlight the importance of AI in the biomedical research and also recent studies that support the use of AI to generate tools using patient data to improve outcomes. Other studies have demonstrated the use of AI to create prediction models to determine response to cancer treatment.

CONCLUSION

The application of AI in the field of pharmaceutical and biomedical studies has been extensive, including cancer research, for diagnosis as well as prognosis of the disease state. It has become a tool for researchers in the management of complex data, ranging from obtaining complementary results to conventional statistical analyses. AI increases the precision in the estimation of treatment effect in cancer patients and determines prediction outcomes.

Models for predicting treatment efficacy of antiepileptic drugs and prognosis of treatment withdrawal in epilepsy patients

Although antiepileptic drugs (AEDs) are the most effective treatment for epilepsy, 30–40% of patients with epilepsy would develop drug-refractory epilepsy. An accurate, preliminary prediction of the efficacy of AEDs has great clinical significance for patient treatment and prognosis. Some studies have developed statistical models and machine-learning algorithms (MLAs) to predict the efficacy of AEDs treatment and the progression of disease after treatment withdrawal, in order to provide assistance for making clinical decisions in the aim of precise, personalized treatment. The field of prediction models with statistical models and MLAs is attracting growing interest and is developing rapidly. What’s more, more and more studies focus on the external validation of the existing model. In this review, we will give a brief overview of recent developments in this discipline.

Clinical Management of Drug Resistant Epilepsy: A Review on Current Strategies

Drug resistant epilepsy (DRE) is defined as the persistence of seizures despite at least two syndrome-adapted antiseizure drugs (ASD) used at efficacious daily dose. Despite the increasing number of available ASD, about a third of patients with epilepsy still suffer from drug resistance. Several factors are associated with the risk of evolution to DRE in patients with newly diagnosed epilepsy, including epilepsy onset in the infancy, intellectual disability, symptomatic epilepsy and abnormal neurological exam. Pharmacological management often consists in ASD polytherapy. However, because quality of life is driven by several factors in patients with DRE, including the tolerability of the treatment, ASD management should try to optimize efficacy while anticipating the risks of drug-related adverse events. All patients with DRE should be evaluated at least once in a tertiary epilepsy center, especially to discuss eligibility for non-pharmacological therapies. This is of paramount importance in patients with drug resistant focal epilepsy in whom epilepsy surgery can result in long-term seizure freedom. Vagus nerve stimulation, deep brain stimulation or cortical stimulation can also improve seizure control. Lastly, considering the effect of DRE on psychologic status and social integration, comprehensive care adaptations are always needed in order to improve patients' quality of life.

Machine learning and wearable devices of the future

Machine learning (ML) is increasingly recognized as a useful tool in healthcare applications, including epilepsy. One of the most important applications of ML in epilepsy is seizure detection and prediction, using wearable devices (WDs). However, not all currently available algorithms implemented in WDs are using ML. In this review, we summarize the state of the art of using WDs and ML in epilepsy, and we outline future development in these domains. There is published evidence for reliable detection of epileptic seizures using implanted electroencephalography (EEG) electrodes and wearable, non-EEG devices. Application of ML using the data recorded with WDs from a large number of patients could change radically the way we diagnose and manage patients with epilepsy.

Artificial Intelligence in Pharmacoepidemiology: A Systematic Review. Part 1-Overview of Knowledge Discovery Techniques in Artificial Intelligence

Aim

To perform a systematic review on the application of artificial intelligence (AI) based knowledge discovery techniques in pharmacoepidemiology.

Study Eligibility Criteria

Clinical trials, meta-analyses, narrative/systematic review, and observational studies using (or mentioning articles using) artificial intelligence techniques were eligible. Articles without a full text available in the English language were excluded.

Data Sources

Articles recorded from 1950/01/01 to 2019/05/06 in Ovid MEDLINE were screened.

Participants

Studies including humans (real or simulated) exposed to a drug.

Results

In total, 72 original articles and 5 reviews were identified via Ovid MEDLINE. Twenty different knowledge discovery methods were identified, mainly from the area of machine learning (66/72; 91.7%). Classification/regression (44/72; 61.1%), classification/regression + model optimization (13/72; 18.0%), and classification/regression + features selection (12/72; 16.7%) were the three most frequent tasks in reviewed literature that machine learning methods has been applied to solve. The top three used techniques were artificial neural networks, random forest, and support vector machines models.

Conclusions

The use of knowledge discovery techniques of artificial intelligence techniques has increased exponentially over the years covering numerous sub-topics of pharmacoepidemiology.

Systematic Review Registration

Systematic review registration number in PROSPERO: CRD42019136552.

The Differences Between Individuals Engaging in Nonsuicidal Self-Injury and Suicide Attempt Are Complex (vs. Complicated or Simple)

BACKGROUND

Why do some people engage in nonsuicidal self-injury (NSSI) while others attempt suicide? One way to advance knowledge about this question is to shed light on the differences between people who engage in NSSI and people who attempt suicide. These groups could differ in three broad ways. First, these two groups may differ in a simple way, such that one or a small set of factors is both necessary and sufficient to accurately distinguish the two groups. Second, they might differ in a complicated way, meaning that a specific set of a large number of factors is both necessary and sufficient to accurately classify them. Third, they might differ in a complex way, with no necessary factor combinations and potentially no sufficient factor combinations. In this scenario, at the group level, complicated algorithms would either be insufficient (i.e., no complicated algorithm produces good accuracy) or unnecessary (i.e., many complicated algorithms produce good accuracy) to distinguish between groups. This study directly tested these three possibilities in a sample of people with a history of NSSI and/or suicide attempt.

METHOD

A total of 954 participants who have either engaged in NSSI and/or suicide attempt in their lifetime were recruited from online forums. Participants completed a series of measures on factors commonly associated with NSSI and suicide attempt. To test for simple differences, univariate logistic regressions were conducted. One theoretically informed multiple logistic regression model with suicidal desire, capability for suicide, and their interaction term was considered as well. To examine complicated and complex differences, multiple logistic regression and machine learning analyses were conducted.

RESULTS

No simple algorithm (i.e., single factor or small set of factors) accurately distinguished between groups. Complicated algorithms constructed with cross-validation methods produced fair accuracy; complicated algorithms constructed with bootstrap optimism methods produced good accuracy, but multiple different algorithms with this method produced similar results.

CONCLUSIONS

Findings were consistent with complex differences between people who engage in NSSI and suicide attempts. Specific complicated algorithms were either insufficient (cross-validation results) or unnecessary (bootstrap optimism results) to distinguish between these groups with high accuracy.

Machine learning prediction of incidence of Alzheimer's disease using large-scale administrative health data

Nationwide population-based cohort provides a new opportunity to build an automated risk prediction model based on individuals' history of health and healthcare beyond existing risk prediction models. We tested the possibility of machine learning models to predict future incidence of Alzheimer's disease (AD) using large-scale administrative health data. From the Korean National Health Insurance Service database between 2002 and 2010, we obtained de-identified health data in elders above 65 years (N = 40,736) containing 4,894 unique clinical features including ICD-10 codes, medication codes, laboratory values, history of personal and family illness and socio-demographics. To define incident AD we considered two operational definitions: "definite AD" with diagnostic codes and dementia medication (n = 614) and "probable AD" with only diagnosis (n = 2026). We trained and validated random forest, support vector machine and logistic regression to predict incident AD in 1, 2, 3, and 4 subsequent years. For predicting future incidence of AD in balanced samples (bootstrapping), the machine learning models showed reasonable performance in 1-year prediction with AUC of 0.775 and 0.759, based on "definite AD" and "probable AD" outcomes, respectively; in 2-year, 0.730 and 0.693; in 3-year, 0.677 and 0.644; in 4-year, 0.725 and 0.683. The results were similar when the entire (unbalanced) samples were used. Important clinical features selected in logistic regression included hemoglobin level, age and urine protein level. This study may shed a light on the utility of the data-driven machine learning model based on large-scale administrative health data in AD risk prediction, which may enable better selection of individuals at risk for AD in clinical trials or early detection in clinical settings.

Harnessing repeated measurements of predictor variables for clinical risk prediction: a review of existing methods

BACKGROUND

Clinical prediction models (CPMs) predict the risk of health outcomes for individual patients. The majority of existing CPMs only harness cross-sectional patient information. Incorporating repeated measurements, such as those stored in electronic health records, into CPMs may provide an opportunity to enhance their performance. However, the number and complexity of methodological approaches available could make it difficult for researchers to explore this opportunity. Our objective was to review the literature and summarise existing approaches for harnessing repeated measurements of predictor variables in CPMs, primarily to make this field more accessible for applied researchers.

METHODS

MEDLINE, Embase and Web of Science were searched for articles reporting the development of a multivariable CPM for individual-level prediction of future binary or time-to-event outcomes and modelling repeated measurements of at least one predictor. Information was extracted on the following: the methodology used, its specific aim, reported advantages and limitations, and software available to apply the method.

RESULTS

The search revealed 217 relevant articles. Seven methodological frameworks were identified: time-dependent covariate modelling, generalised estimating equations, landmark analysis, two-stage modelling, joint-modelling, trajectory classification and machine learning. Each of these frameworks satisfies at least one of three aims: to better represent the predictor-outcome relationship over time, to infer a covariate value at a pre-specified time and to account for the effect of covariate change.

CONCLUSIONS

The applicability of identified methods depends on the motivation for including longitudinal information and the method's compatibility with the clinical context and available patient data, for both model development and risk estimation in practice.

Involvement of hypoxia-inducible factor-1 alpha in the upregulation of P-glycoprotein in refractory epilepsy

To explore the involvement of hypoxia-inducible factor-1 alpha (HIF-1α) in the upregulation of P-glycoprotein (P-gp) in refractory epilepsy. Brain tissue specimens were collected and analyzed for expression of HIF-1α and P-gp using an immunohistochemical (IHC) staining method in both refractory epilepsy group and control group. Correlation between HIF-1α and P-gp expression level in refractory epilepsy group was analyzed. Then, a hypoxia cell model was established by simulating the nerve cell hypoxic microenvironment in the human U251 cell line using cobalt chloride (CoCl2). Western blot analysis was used to detect expression levels of HIF-1α and P-gp in the hypoxic cell model. Finally, expression of HIF-1α and P-gp was detected using real-time quantitative PCR and Western blot, respectively, after U251 hypoxic model cells were infected with HIF-1α siRNA. IHC scores of HIF-1α and P-gp in refractory epilepsy group were significantly higher than that in control group. In addition, the expression of HIF-1α was positively correlated with the expression of P-gp in refractory epilepsy group. Expression levels of HIF-1α and P-gp in U251 cells cultured with 250 µmol/L CoCl2 for 48 hours were significantly higher than that in controls. After transfection with siRNA targeting HIF-1α, expressions of HIF-1α and P-gp at mRNA and protein level were decreased, respectively, in the hypoxia cell model. HIF-1α may be involved in the upregulation of P-gp in refractory epilepsy through inducement of P-gp expression. Therefore, activation of the HIF-1α/P-gp pathway is one hypothesis proposed to explain the pathogenesis of refractory epilepsy.

Machine learning applications in epilepsy

Machine learning leverages statistical and computer science principles to develop algorithms capable of improving performance through interpretation of data rather than through explicit instructions. Alongside widespread use in image recognition, language processing, and data mining, machine learning techniques have received increasing attention in medical applications, ranging from automated imaging analysis to disease forecasting. This review examines the parallel progress made in epilepsy, highlighting applications in automated seizure detection from electroencephalography (EEG), video, and kinetic data, automated imaging analysis and pre-surgical planning, prediction of medication response, and prediction of medical and surgical outcomes using a wide variety of data sources. A brief overview of commonly used machine learning approaches, as well as challenges in further application of machine learning techniques in epilepsy, is also presented. With increasing computational capabilities, availability of effective machine learning algorithms, and accumulation of larger datasets, clinicians and researchers will increasingly benefit from familiarity with these techniques and the significant progress already made in their application in epilepsy.

A scale for prediction of response to AEDs in patients with MRI-negative epilepsy

OBJECTIVES

Antiepileptic drugs (AEDs) are the first choice in magnetic resonance imaging (MRI)-negative patients with epilepsy, although the responses to AEDs are diverse. Preoperative evaluation and postoperative prognosis in MRI-negative epilepsy have been reported. However, there are few tools for predicting the response to AEDs. Herein, we developed an AED response scale based on clinical factors and video-electroencephalography (VEEG) in MRI-negative patients with epilepsy.

METHODS

A total of 132 consecutive patients with MRI-negative epilepsy at the Epilepsy Center of Henan Provincial People's Hospital between August 2016 and August 2018 were included. Patients were further divided into drug-responsive epilepsy ([DSE-MRI (-)]; n = 101) and drug-resistant epilepsy ([DRE-MRI (-)]; n = 31) groups. The clinical and VEEG factors were evaluated in univariate analyses and multivariate logistic regression analyses. A scale was derived and the scores categorized into 3 risk levels of DRE-MRI (-).

RESULTS

A scale was established based on 4 independent risk factors for DRE-MRI (-). The scale had a sensitivity of 83.87%, specificity of 80.20%, positive likelihood ratio of 4.24, negative likelihood ratio of 0.20, and showed good discrimination with the area under the curve (AUC) of 0.886 (0.826-0.946). The categorization of the risk score based on this scale was: low risk (0-3 points), medium risk (3-5 points), and high risk (>5 points).

CONCLUSION

We established a DRE-MRI (-) scale with a good sensitivity and specificity, which may be useful for clinicians when making medical decisions in patients with MRI-negative epilepsy.