Wavelet Energy and Wavelet Coherence as EEG Biomarkers for the Diagnosis of Parkinson’s Disease-Related Dementia and Alzheimer’s Disease

Abstract

Abstract:Parkinson's disease (PD) and Alzheimer's disease (AD) can coexist in severely affected; elderly patients. Since they have different pathological causes and lesions and consequently require different treatments; it is critical to distinguish PD-related dementia (PD-D) from AD. Conventional electroencephalograph (EEG) analysis has produced poor results. This study investigated the possibility of using relative wavelet energy (RWE) and wavelet coherence (WC) analysis to distinguish between PD-D patients; AD patients and healthy elderly subjects. In EEG signals; we found that low-frequency wavelet energy increased and high-frequency wavelet energy decreased in PD-D patients and AD patients relative to healthy subjects. This result suggests that cognitive decline in both diseases is potentially related to slow EEG activity; which is consistent with previous studies. More importantly; WC values were lower in AD patients and higher in PD-D patients compared with healthy subjects. In particular; AD patients exhibited decreased WC primarily in the γ band and in links related to frontal regions; while PD-D patients exhibited increased WC primarily in the α and β bands and in temporo-parietal links. Linear discriminant analysis (LDA) of RWE produced a maximum accuracy of 79.18% for diagnosing PD-D and 81.25% for diagnosing AD. The discriminant accuracy was 73.40% with 78.78% sensitivity and 69.47% specificity. In distinguishing between the two diseases; the maximum performance of LDA using WC was 80.19%. We suggest that using a wavelet approach to evaluate EEG results may facilitate discrimination between PD-D and AD. In particular; RWE is useful for differentiating individuals with and without dementia and WC is useful for differentiating between PD-D and AD.

Systematic Review on Resting-State EEG for Alzheimer’s Disease Diagnosis and Progression Assessment

“…In those studies, distinctions in spectral slowing features between AD and other dementias are identified for vascular dementia [ 26 , 39 , 130 ], frontotemporal dementia (FTD) [ 37 , 130 , 175 , 176 ], Lewy body dementia (DLB) [ 130 , 176 ], and Parkison's disease dementia (PDD) [ 130 ]. Moreover, disparity in synchronization measures is reported between AD, PDD [ 38 , 42 ], DLB and FTD [ 42 ]. Additionally, a combined model of EEG and MRI improved discrimination between AD and DLB [ 41 ].…”

“…The most utilized layout is the international 10-20 system, commonly consisting of 21 electrodes; higher density variants of the 10-20 system such as 10-10 and 10-5 systems are utilized as well, usually with 64 and 128 electrodes, respectively, [ 30 ] as well as the alternative layouts Maudsley [ 31 ] and Geodesics positioning systems [ 32 ]. In recent years, quantitative EEG (qEEG, henceforth simply EEG) has been proven to be a reliable clinical tool for the diagnosis and study of illnesses and cortical disorders such as Huntington disease [ 33 ], autism spectrum disorders [ 34 ], epilepsy and seizure [ 35 ], cerebral ischemia [ 36 ], frontotemporal dementia [ 37 ], and Parkinson's disease dementia [ 38 ]. Furthermore, the differential diagnosis between AD and other diseases that lead to dementia as vascular brain injury [ 39 , 40 ] and Lewy body diseases [ 41 , 42 ] was assessed with EEG.…”

Functional dynamics of thalamic local field potentials correlate with modulation of neuropathic pain

“…In the DNSI approach, oscillations in LFPs were represented with WP coefficients through a WP transform. Oscillations with high amplitude and regular patterns were assigned larger WP coefficients, while those with low amplitude and random patterns were assigned smaller WP coefficients (Donoho & Johnstone, ; Jeong, Kim, Song, Chung, & Jeong, ; Luo et al., ). A state discrimination model with a statistically optimal adaptive threshold and discrimination strategy was used to classify the coefficients into two classes.…”

Topographical pattern analysis using wavelet based coherence connectivity estimation in the distinction of meditation and non-meditation EEG

“…It is the time-frequencyspace by successively convolving the time series with scaled and translated versions of a wavelet function. This waveletbased coherence measurement is deliberately significant in the contribution of the functional brain connectivity and, using specifications that anticipate the brain activity involved in a particular cognitive action [24].…”