Early identification of seizure freedom with medical treatment in patients with mesial temporal lobe epilepsy and hippocampal sclerosis

Hippocampal sclerosis: A review on current research status and its mechanisms

Hippocampal sclerosis (HS) is a pathological condition characterized by significant loss of hippocampal neurons and gliosis. This condition represents the most common neuropathological change observed in patients with temporal lobe epilepsy (TLE) and is also found in aging individuals. TLE related to HS is the most prevalent type of drug-resistant epilepsy in adults, and its underlying mechanisms are not yet fully understood. Therefore, developing improved methods for predicting and treating drug-resistant patients with TLE-HS is crucial. Patients with TLE-HS often experience cognitive impairment and psychological comorbidities, significantly affecting their quality of life. Consequently, a thorough review of the current research status of TLE-HS is essential, focusing on its prediction, diagnosis, treatment, and underlying mechanisms. The hippocampus plays a pivotal role in memory and cognition. HS of aging (HS-Aging), a condition linked to dementia in the ultra-elderly, is marked by severe CA1 (cornu ammonis) neuronal loss and frequent transactive response DNA-binding protein of 43 kDa (TDP-43) proteinopathy, often misdiagnosed as Alzheimer's disease (AD). Nonetheless, clinical characteristics and patterns of hippocampal atrophy can help differentiate between the two disorders. This review aims to provide a comprehensive overview of the pathological features of HS, the relevant mechanisms underlying TLE-HS and HS-Aging, current imaging diagnostic techniques, including machine learning, and available treatment modalities. It also explores the prognosis and comorbidities related to these conditions. Future research directions include establishing animal models to clarify the poorly understood mechanisms underlying HS, particularly those related to emotional processing. Investigating post-HS behavioral and cognitive changes in these models will lay the foundation for further advancements in this field. This review is a cornerstone for future investigations and suggests additional research endeavors.

How accurate are machine learning models in predicting anti-seizure medication responses: A systematic review

IMPORTANCE

Current epilepsy management protocols often depend on anti-seizure medication (ASM) trials and assessment of clinical response. This may delay the initiation of the ASM regimen that might optimally balance efficacy and tolerability for individual patients. Machine learning (ML) can offer a promising tool for efficiently predicting ASM response.

OBJECTIVE

The objective of this review is to synthesize the available information about the effectiveness and limitations of ML models in predicting and classifying the response of patients with epilepsy to ASMs, and to assess the impact of various data inputs on prediction performance.

EVIDENCE REVIEW

We conducted a comprehensive search of studies utilizing ML models for ASM response prediction using PubMed and Scopus up until November 2024.

FINDINGS

The review included 37 studies. Various data types, including clinical information, brain MRI, EEG, and genetic data, are useful in predicting responses to ASMs. Tree-based ML algorithms and Support Vector Machines are the most used models. Reported results vary widely, with certain models achieving near-perfect accuracy and others performing similar to random classifiers. The review also highlights the limitations of this research field, especially concerning the quality and quantity of data.

CONCLUSIONS AND RELEVANCE

The findings indicate that while ML models show great promise in predicting ASM responses in epilepsy, further research is required to refine these models for practical clinical application. The review underscores both the potential of ML in advancing precision medicine in epilepsy management and the need for continued research to improve prediction accuracy.

Machine learning models for predicting treatment response in infantile epilepsies

Epilepsy stands as one of the prevalent and significant neurological disorders, representing a critical healthcare challenge. Recently, machine learning techniques have emerged as versatile tools across various healthcare domains, encompassing diagnostics, treatment assessment, and prognosis. We compared 11 machine learning model to find the best ML model to predict drug treatment outcomes for our cohort, which we previously evaluated using classical statistical methods.

METHODS

In our study, we evaluated patients who presented to the pediatric neurology department of our university hospital with seizures at the age of 1 to 24 months and were diagnosed with epilepsy. We utilized 11 different machine learning techniques namely Decision Tree, Bagging, K-Nearest Neighbour, Linear Discriminant Analysis, Logistic Regression, Neural Networks, Deep Neural Networks, Support Vector Machine. Besides, we compared these techniques using various performance metrics to identify anti-seizure medicine response. We also utilized the chi-square feature selection methods to enhance performance in machine learning algorithms.

RESULTS

Two hundred and twenty-nine patients (110 male and 119 female) who were diagnosed between the ages of 1-24 months were included in the study. Support Vector Machine algorithm was found to be effective in drug resistant epilepsy detection, with the highest aure under curve value (0.9934) and achieving a test accuracy of 97.06 %.

CONCLUSION

This study can shed light on future studies by showing that the Support Vector Machine algorithm can effectively determine the drug resistant epilepsy. The pediatric neurologist and experts should be referred to non-medical treatment (epilepsy surgery, ketogenic diet) at the early stages and multidisciplinary approach should be provided.

Long-term outcome of treatment-naïve patients with mesial temporal lobe epilepsy with hippocampal sclerosis: A retrospective study in a single center

PURPOSE

This study aimed to describe long-term treatment outcomes of treatment-naïve patients with mesial temporal lobe epilepsy with hippocampal sclerosis (MTLE-HS).

METHODS

A retrospective review was conducted of treatment-naïve patients with MTLE-HS who visited the Yonsei Epilepsy Clinic from April 2000 to April 2022 and were followed up for at least 2 years. Seizure freedom (SF) was defined as no seizures or auras only for >1 year, and complete SF was defined as no seizures including auras for >1 year.

RESULTS

Eighty-four treatment-naïve patients with MTLE-HS with a median follow-up of 122 months were included. Except for one patient who underwent early surgical treatment, of the remaining 83 patients, 31 (37.3 %) achieved SF and remained in remission, 38 (45.8 %) had fluctuations in seizure control, and 14 (16.9 %) never achieved SF. Additionally, 18 (21.7 %) patients achieved complete SF and remained in remission, 42 (50.6 %) showed fluctuations, and 23 (27.7 %) never achieved complete SF. Fifty-three (63.9 %) patients achieved SF and 34 (41.0 %) achieved complete SF at their last visit. Older age at epilepsy onset, male sex, low pretreatment seizure density, history of central nervous system infection before age 5, absence of aura, and fewer antiseizure medications in the final regimen were associated with favorable outcome. Of the 84 patients, 11 (13.1 %) underwent temporal lobectomy.

CONCLUSIONS

Medical treatment outcomes in treatment-naïve MTLE-HS were relatively better than previously reported outcomes in MTLE-HS, although frequent fluctuations in seizure control were observed.

Machine Learning and Artificial Intelligence Applications to Epilepsy: a Review for the Practicing Epileptologist

PURPOSE OF REVIEW

Machine Learning (ML) and Artificial Intelligence (AI) are data-driven techniques to translate raw data into applicable and interpretable insights that can assist in clinical decision making. Some of these tools have extremely promising initial results, earning both great excitement and creating hype. This non-technical article reviews recent developments in ML/AI in epilepsy to assist the current practicing epileptologist in understanding both the benefits and limitations of integrating ML/AI tools into their clinical practice.

RECENT FINDINGS

ML/AI tools have been developed to assist clinicians in almost every clinical decision including (1) predicting future epilepsy in people at risk, (2) detecting and monitoring for seizures, (3) differentiating epilepsy from mimics, (4) using data to improve neuroanatomic localization and lateralization, and (5) tracking and predicting response to medical and surgical treatments. We also discuss practical, ethical, and equity considerations in the development and application of ML/AI tools including chatbots based on Large Language Models (e.g., ChatGPT). ML/AI tools will change how clinical medicine is practiced, but, with rare exceptions, the transferability to other centers, effectiveness, and safety of these approaches have not yet been established rigorously. In the future, ML/AI will not replace epileptologists, but epileptologists with ML/AI will replace epileptologists without ML/AI.