Analyzing Phenotypes Using Artificial Intelligence Tools in Subgroups of Absence and Myoclonic Epilepsies

Neurology(2018)

引用 23|浏览25
暂无评分
摘要
Objective: To reanalyze seizure phenotypes in families with childhood absence epilepsy (CAE) and juvenile myoclonic epilepsy (JME) using artificial intelligence tools and results of Whole Exome/Genome Sequencing (WES/WGS). Background: Since 1992, the Genetic Epilepsy StudieS Consortium (GENESS) and collaborators have collected phenotype data on CAE and JME families from Los Angeles, CA (USA), Mexico, Honduras, El Salvador, Guatemala, Peru, Brazil, Spain, France, and Japan. Combining new informatics tools, the new 2016 ILAE classification of seizures and epilepsies and WES/WGS can help further reanalysis of cases. Design/Methods: The phenotypes of 268 probands with JME were reanalyzed using the RapidMiner tool. We studied the variables age, gender, seizure types, age at onset of each type, trigger factors, EEG findings, family history, and response to treatment. We used the pediatric age groups suggested by the National Institutes of Child Health and Human Development. Results: The system organized probands in 8 clusters. Clusters with more clinical sense happened when seizure type and age at onset were the main criteria. Some examples are: Cluster 1= early childhood absence (AS) with-without eyelid myoclonia (EM) plus early adolescence mioclonias (MS) and tonic-clonic seizures (TC), Cluster 2= early AS with early adolescence MS and TC with-without FS or EM, Cluster 3= middle childhood AS with middle childhood MS and TC without EM; and Cluster 4= Noise or no specific pattern. Cluster analysis results were then correlated with recently discovered epilepsy genes: EFHC1, ICK, IPO8, PROSER1 and MYOFERLIN. Conclusions: Some clusters found match the diagnosis in the new 2016 and old 1989 epilepsy classifications, but some clusters together with the new epilepsy genes suggest endophenotypes or new subsyndromes such as (1) middle childhood absence with middle childhood myoclonias associated with EFHC1, PROSER1 and ICK variants; (2) early AS with adolescent MS, EM, TC associated with IPO8 and MYOFERLIN variants. Study Supported by: National Institutes of Health (1R01NS055057), VA Merit Review (5I01CX000743) and UNITEC. Disclosure: Dr. Duron has received personal compensation for consulting, serving on a scientific advisory board, speaking, or other activities with Abbot. Dr. Arias has nothing to disclose. Dr. Martinez has nothing to disclose. Dr. Patterson has nothing to disclose. Dr. Bailey has nothing to disclose. Dr. Tanaka has nothing to disclose. Dr. Ochoa has nothing to disclose. Dr. Jara-Prado has nothing to disclose. Dr. Alonso has nothing to disclose. Dr. Medina has nothing to disclose. Dr. Nguyen has nothing to disclose. Dr. Delgado-Escueta has nothing to disclose.
更多
查看译文
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要