Clinical usefulness and feasibility of time-frequency analysis of chemosensory event-related potentials

Background: The clinical usefulness of olfactory event-related brain potentials (OERPs) to assess olfactory function is limited by the relatively low signal-to-noise ratio of the responses identified using conventional time-domain averaging. Recently, it was shown that time-frequency analysis of the obtained EEG signals can markedly improve the signal-to-noise ratio of OERPs in healthy controls, because it enhances both phase-locked and non phase-locked EEG responses. The aim of the present study was to investigate the clinical usefulness of this approach and evaluate its feasibility in a clinical setting. Methodology: We retrospectively analysed EEG recordings obtained from 45 patients (15 anosmic, 15 hyposmic and 15 normos- mic). The responses to olfactory stimulation were analysed using conventional time-domain analysis and joint time-frequency analysis. The ability of the two methods to discriminate between anosmic, hyposmic and normosmic patients was assessed using a Receiver Operating Characteristic analysis. Results: The discrimination performance of OERPs identified using conventional time-domain averaging was poor. In contrast, the discrimination performance of the EEG response identified in the time-frequency domain was relatively high. Furthermore, we found a significant correlation between the magnitude of this response and the psychophysical olfactory score. Conclusion: Time-frequency analysis of the EEG responses to olfactory stimulation could be used as an effective and reliable diagnostic tool for the objective clinical evaluation of olfactory function in patients.

Download Full-text

TIME-FREQUENCY ANALYSIS OF MISMATCH NEGATIVITY IN SCHIZOPHRENIA PATIENTS IN TAIWAN

Biomedical Engineering Applications Basis and Communications ◽

10.4015/s1016237211002645 ◽

2011 ◽

Vol 23 (04) ◽

pp. 287-293

Author(s):

Jung-Lung Hsu ◽

Ming H. Hsieh ◽

Yi-Li Tseng ◽

Ming-Jang Chiu ◽

Chih-Min Liu ◽

...

Keyword(s):

Information Processing ◽

Frequency Analysis ◽

Mismatch Negativity ◽

Phase Coherence ◽

Network Activity ◽

Oddball Paradigm ◽

Time Frequency Analysis ◽

Time Frequency ◽

Grand Average ◽

Related Potentials

Background: Event-related potentials (ERPs) are currently utilized to study information processing under the influence of neuropsychiatric disorders and are candidate endophenotypes of schizophrenia. However, the traditional ERP approach discards most oscillatory information in the EEG through grand averaging process. This study tried to focus on ERP component of mismatch negativity (MMN) between control and schizophrenia groups in Taiwan. In addition to traditional (grand average) ERP approach, we also compared the time-frequency analysis results of the mismatch negativity. Methods: MMN values were recorded in 55 schizophrenia patients and 42 controls using an auditory oddball paradigm for the duration of the MMN. Besides the traditional grand average approach, the continuous data files were also used for time-frequency analysis via ERP wavelab. Six parameters were explored: event-related spectral perturbation, WTav, avWT, induced activity, intertrial phase coherence and intertrial linear coherence. Results: Traditional ERP approach discovered that group difference occurred in MMN mean amplitude which was compatible with the previous literature. In time-frequency analysis, schizophrenia subjects have lower intertrial phase coherence than controls. Conclusion: Deficit in preattentive sensory information processing, such as MMN, is prominent in Taiwanese schizophrenia subjects. It may also play a role in the basic integrated neural network activity, which could be further explored by time-frequency approach.

Download Full-text