Assessment of classification improvement in patients with Alzheimer's disease based on magnetoencephalogram blind source separation

Abstract

2 SummaryObjectives: In this pilot study, we intended to assess whether a procedure based on blind source separation (BSS) and subsequent partial reconstruction of magnetoencephalogram (MEG) recordings might enhance the differences between MEGs from Alzheimer's disease (AD) patients and elderly control subjects. Materials and methods:We analysed MEG background activity recordings acquired with a 148-channel whole-head magnetometer from 21 AD patients and 21 control subjects. Artefact-free epochs of 20 s were blindly decomposed using the algorithm for multiple unknown signals extraction (AMUSE), which arranges the extracted components by decreasing linear predictability. Thus, the components of diverse epochs and subjects could be easily compared. Every component was characterised with its median frequency and spectral entropy (denoted by f median and SpecEn, respectively). The differences between subject groups in these variables were statistically evaluated to find out which components could improve the subject classification. Then, these significant components were used to partially reconstruct the MEG recordings.Results: The statistical analysis showed that the AMUSE components which provided the largest differences between demented patients and control subjects were ordered together. Considering this analysis, we defined two subsets, denoted by BSS-{15,35} and BSS-{20,30}, which included 21 components (15 to 35) and 11 components (20 to 30), respectively. We 3 partially reconstructed the MEGs with these subsets. Then, the classification performance was computed with a leave-one-out cross-validation procedure for the case where no BSS was applied and for the partial reconstructions BSS-{15,35} and BSS-{20,30}. The BSS and component selection procedure improved the classification accuracy from 69.05% to 83.33% using f median with BSS-{15,35} and from 61.91% to 73.81% using SpecEn with BSS-{20,30}. Conclusion:These preliminary results lead us to think that the proposed procedure based on BSS and selection of significant components may improve the classification of AD patients using straightforward features from MEG recordings.

Blind source separation to enhance spectral and non-linear features of magnetoencephalogram recordings. Application to Alzheimer's disease

“…Additionally, we aimed at determining whether the range of BSS components with significant differences between demented patients and controls differed when both kinds of features (spectral and non-linear ones) were considered. We also intended to confirm the results of a previous pilot study [14].…”

“…Moreover, considering m = n allows us to consistently compare the same number of extracted components instead of estimating different values of n for each signal epoch [14].…”

“…It has been shown that some BSS components may be more affected by AD than others [12,14]. Therefore, the range of the most sensitive components to AD -y range (t) -may be back projected using the inverse of W (W -1 ) to compute a partial reconstruction of the MEGs -x partial (t) -that may have enhanced features of AD:…”

“…Considering these research studies, it can be hypothesised that the application of BBS, together with features extracted from electromagnetic brain activity recordings, may enhance features associated with diseases like AD. This is due to the fact that some BSS components of the EEG and MEG signals may be more sensitive to AD than others [12,14,22]. Hence, the most relevant components may be selected and the electromagnetic brain signals may be partially reconstructed using only these components to achieve a better discrimination between AD patients and healthy subjects [14].…”

“…Nevertheless, it is desirable to develop novel strategies to help in AD detection from the analysis of the electromagnetic brain activity [9,11,12]. Techniques based on spatial filtering can help to achieve this goal, as these algorithms offer additional perspectives to examine EEG and MEG signals [11][12][13][14]. For instance, common spatial patterns (CSP) have been recently applied to enhance characteristics of EEG recordings in mild cognitive impairment (MCI) patients who eventually developed AD [11].…”