- 【Updated on May 12, 2025】 Integration of CiNii Dissertations and CiNii Books into CiNii Research
- Trial version of CiNii Research Knowledge Graph Search feature is available on CiNii Labs
- 【Updated on June 30, 2025】Suspension and deletion of data provided by Nikkei BP
- Regarding the recording of “Research Data” and “Evidence Data”
Adaptive Abnormality Detection on ECG Signal by Utilizing FLAC Features
Description
In this paper we propose a self-adaptive algorithm for noise robust abnormality detection on ECG data. For extracting features from ECG signals, we propose a feature extraction method by characterizing the magnitude, frequency and phase information of ECG signal as well as the temporal dynamics in time and frequency domains. At abnormality detection stage, we employ the subspace method for adaptively modeling the principal pattern subspace of ECG signal in unsupervised manner. Then, we measure the dissimilarity between the test signal and the trained major pattern subspace. The atypical periods can be effectively discerned based on such dissimilarity degree. The experimental results validate the effectiveness of the proposed approach for mining abnormalities of ECG signal including promising performance, high efficiency and robust to noise.