Separation of data sources and extraction of target components of the electrophysiological time series on the example of fetal ECG
Electrophysiological time series are a valuable source of information but are often presented by instantaneous unknown linear mixtures of sources. Thus, one faced the task of data extraction, canceling other sources, to identify target data components. Fetal electrocardiography is one of the most im...
Збережено в:
| Видавець: | Інститут проблем реєстрації інформації НАН України |
|---|---|
| Дата: | 2023 |
| Автори: | , |
| Формат: | Стаття |
| Мова: | Ukrainian |
| Опубліковано: |
Інститут проблем реєстрації інформації НАН України
2023
|
| Теми: | |
| Онлайн доступ: | http://drsp.ipri.kiev.ua/article/view/287017 |
| Теги: |
Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
|
Організація
Data Recording, Storage & Processing| Резюме: | Electrophysiological time series are a valuable source of information but are often presented by instantaneous unknown linear mixtures of sources. Thus, one faced the task of data extraction, canceling other sources, to identify target data components. Fetal electrocardiography is one of the most important and valuable sources of information. However, the resulting abdominal electrophysiological data is a mixed source of maternal and fetal electrocardiograms, and also muscle activity data, noises, etc. The abdominal electrocardiogram preprocessing, noise removing, cancellation of the maternal electrocardiogram while maintaining the diagnostically valuable components of the fetal electrocardiogram is the most important steps in the fetal electrocardiogram extraction from the abdominal electrocardiogram. Analyzing source separation methods made it possible to highlight the advantages and disadvantages of current techniques, and to form a sequence for electrophysiological data processing. The proposed approach consists of four stages. In the first stage, electrophysiological data are preprocessed, including the baseline canceling and removing high-frequency noise using a second-order low-pass filter with zero phase. However, decorrelation based on principal component analysis is applied. The next step based on using a source separation algorithm to target data extraction. In the fourth step, channels of non-target data are canceled through a generalized eigenvalue decomposition and periodic components analysis. Thus, components of the target fetal electrocardiogram data are extracted from the mixed data. The experiment was carried out using a record that consists of five abdominal channels and one chest lead which corresponded to maternal electrocardiogram. The results of the standard error and the signal-to-noise ratio show that the presented approach is appropriate for extracting the fetal electrocardiogram from the abdominal channels and the following analysis. |
|---|