CHB-MIT (CHB-MIT Scalp EEG)

CHB-MIT 数据集是来自患有顽固性癫痫发作的儿科受试者的脑电图记录数据集。在停用抗癫痫药物后，对受试者进行长达数天的监测，以表征他们的癫痫发作情况并评估他们是否适合进行手术干预。该数据集包含 23 名患者，分为 24 个病例（一名患者有 2 个记录，相隔 1.5 年）。该数据集包含 969 小时的头皮脑电图记录，其中 173 次癫痫发作。数据集中存在各种类型的癫痫发作（阵挛、失张力、强直）。数据集中包含的患者（男性、女性、10-22 岁）和不同类型的癫痫发作的多样性是评估自动癫痫发作检测方法在现实环境中的性能的理想选择。

The CHB-MIT dataset is a collection of scalp electroencephalogram (EEG) recordings from pediatric subjects with refractory epileptic seizures. Subjects were monitored for up to several days after discontinuing antiepileptic drugs to characterize their seizure phenotypes and assess their eligibility for surgical intervention. The dataset includes 23 patients divided into 24 cases, with one patient having two recordings separated by 1.5 years. It contains 969 hours of scalp EEG recordings encompassing 173 epileptic seizures. Various types of epileptic seizures are present in the dataset, including clonic, atonic, and tonic seizures. The diversity of the enrolled subjects (both males and females aged 10 to 22 years) and the range of epileptic seizure types make this dataset an ideal resource for evaluating the performance of automated seizure detection methods in real-world clinical settings.

搜集汇总

背景与挑战

背景概述

CHB-MIT数据集是一个头皮脑电图记录数据集，专门针对患有顽固性癫痫发作的儿科患者，包含23名患者（24个病例）的969小时记录和173次癫痫发作事件。该数据集涵盖了多种癫痫发作类型（如阵挛、失张力、强直）以及不同性别和年龄（10-22岁）的患者，具有较高的多样性和现实代表性，非常适合用于开发和评估自动癫痫发作检测算法。