التفاصيل البيبلوغرافية
| العنوان: |
Statistical analysis of electroencephalograms: independent component analysis of event-related potentials |
| المؤلفون: |
Bugli, Céline, Lambert, Philippe, Boulanger, Bruno, Ledent, Edouard, Pereira, Alvaro, Nardone, Pasquale |
| المساهمون: |
UCL - EUEN/STAT - Institut de statistique, Eli Lilly & Company, ULB - Département de Physique |
| سنة النشر: |
2004 |
| المجموعة: |
DIAL@UCL (Université catholique de Louvain) |
| مصطلحات موضوعية: |
EEG, ERP(P300), ICA, Cleaning, Reduction of dimension, Treatment effect |
| الوصف: |
Electroencephalogram (EEG) is an important diagnostic tool in clinical neurophysiology. However, EEGs are not often used in clinical studies because of intrinsic problem like the huge quantity of data or artifacts. In this paper, we shall describe statistical tools to detect and quantify the effect of drugs on the brain by the analysis of EEGs. We first use Independent Component Analysis (ICA) to detect and remove automatically artifacts from EEGs. In the second step, ICA reduces the dimension of the problem. Using data from a clinical trial, we show that eight ICA components can reconstruct more than 80 percents of the data from the twenty-eight electrodes. Some of these eight ICA components can reconstruct an interesting characteristic of the signals (an event-related potential named P300). Finally, we shall show how the analysis of these two components allow to detect and quantify a treatment effect. Lorazepam decreases the P300 peak amplitude and increases the time of occurrence of the P300 peak. |
| نوع الوثيقة: |
report |
| اللغة: |
English |
| العلاقة: |
boreal:91277; http://hdl.handle.net/2078.1/91277Test |
| الإتاحة: |
http://hdl.handle.net/2078.1/91277Test |
| حقوق: |
info:eu-repo/semantics/openAccess |
| رقم الانضمام: |
edsbas.3F07ED8 |
| قاعدة البيانات: |
BASE |
