Poster No:
210
Submission Type:
Abstract Submission
Authors:
Gangyoung Lee1, Ukeob Park1, Byoung Seok Ye2, Seung Wan Kang1,3
Institutions:
1iMediSync, Inc., Seoul, Korea, Republic of, 2Yonsei University College of Medicine, Seoul, Korea, Republic of, 3Data Center for Korean EEG, Seoul National University College of Nursing, Seoul, Korea, Republic of
First Author:
Co-Author(s):
Ukeob Park
iMediSync, Inc.
Seoul, Korea, Republic of
Byoung Seok Ye
Yonsei University College of Medicine
Seoul, Korea, Republic of
Seung Wan Kang
iMediSync, Inc.|Data Center for Korean EEG, Seoul National University College of Nursing
Seoul, Korea, Republic of|Seoul, Korea, Republic of
Introduction:
In recent research, Electroencephalography (EEG) has been established as a powerful biomarker, already in use for predicting both Alzheimer's Disease (AD) and Lewy Body Disease (LBD) [1]. Expanding beyond EEG, Electrocardiography (ECG) also holds potential as a robust biomarker, particularly considering the concept of Brain-Heart Connectivity [2].
Heart Rate Variability (HRV), derived from ECG measurements, serves as a well-established metric reflecting responses to internal physiological states and external stimuli [3][4]. The combined EEG and HRV-based multimodal biometric approach have been noted for its contribution to garnering interest across various studies [5] and enhancing performance [6][7]. Nevertheless, a distinct gap persists in research on Multimodal Biometrics, specifically leveraging the combined potential of EEG and ECG for the classification of AD and LBD.
To address this gap, this study aims to explore and compare HRV specifically between pure AD and pure LBD groups. Through this comparison, we seek to uncover potential multimodal biometric indicators. This exploration will allow us to further investigate and expand the understanding of Brain-Heart Connectivity, utilizing insights derived from both EEG and ECG.
Methods:
Participant
In this study, Electrocardiography (ECG) was measured for individuals in a resting state, including pure Alzheimer's Disease (AD) patients (n=63) and pure Lewy Body Disease (LBD) patients (n=142). Clinical labeling was conducted by experienced specialists at Severance Hospital, Yonsei University, with substantial expertise in the field.
Data & pre-Processing
ECG signals were preprocessed using a 60Hz notch filter to eliminate power line noise, and measurements were taken for a minimum of 3 minutes at sampling rates of 200Hz or 512Hz. To pinpoint accurate R-peaks, a Butterworth bandpass filter (5–15Hz) was applied for preprocessing, and visual inspections ensured signal integrity post-filtering. The Pan-Tompkins algorithm was then employed to compute RR intervals. Data points outside the normal RR interval range (300ms–2000ms) and those deviating by more than twice the standard deviation was removed. Remaining data was interpolated using the cubic spline method.
HRV
Using the R-peaks as the foundation, several metrics were derived from the RR intervals. Time domain features, including SDNN and frequency domain features extracted via FFT analysis-absolute (abs) and log(ln) power-were computed based on these intervals. These derived indices comprised TF (total power), UVLF(~0.0033Hz), VLF(0.0033–0.04Hz), LF(0.04–0.15Hz), HF(0.15–0.4Hz), PNS(parasympathetic nervous system activity, HF/HF+LF), and SNS(sympathetic nervous system activity, LF/HF+LF), resulting in a total of 16 indices.
Analysis
Since both the pure AD and pure LBD groups exceed 30 individuals, it was assumed that they followed a normal distribution. Therefore, an independent samples t-test method was employed to compare each HRV feature between the two groups.
Results:
Sympathetic nervous system indicators, including SNS (p=0.0173) and ln_VLF (p=0.0082), revealed higher activity in pure AD compared to pure LBD. Additionally, UVLF (p=0.0347) recognized as a stress recovery metric, was also found among these indicators. Conversely, the parasympathetic nervous system indicator, PNS(p=0.0173), exhibited higher activity in pure LBD when compared to pure AD. However, no significant differences were observed between the two groups concerning autonomic nervous system activity indicators (TF, SDNN).
Conclusions:
this study confirmed the potential for HRV, in conjunction with EEG, to be utilized as biomarkers for classifying pure AD and pure LBD. Furthermore, it hints at the advantage of using both biomarkers together in the more accurate classification of pure AD and pure LBD.
Disorders of the Nervous System:
Neurodegenerative/ Late Life (eg. Parkinson’s, Alzheimer’s) 1
Neuroanatomy, Physiology, Metabolism and Neurotransmission:
Neuroanatomy Other 2
Keywords:
Degenerative Disease
Limbic Systems
Neurological
Other - Brain-Heart Connectivity
1|2Indicates the priority used for review
Provide references using author date format
[1] Schumacher, J., Thomas, A. J., Peraza, L. R., Firbank, M., Cromarty, R., Hamilton, C. A., ... & Taylor, J. P. (2020). EEG alpha reactivity and cholinergic system integrity in Lewy body dementia and Alzheimer’s disease. Alzheimer's research & therapy, 12(1), 1-12.
[2] de la Cruz, F., Schumann, A., Köhler, S., Reichenbach, J. R., Wagner, G., & Bär, K. J. (2019). The relationship between heart rate and functional connectivity of brain regions involved in autonomic control. Neuroimage, 196, 318-328.
[3] Ernst, G. (2017). Heart-rate variability—more than heart beats?. Frontiers in public health, 5, 240.
[4] Porges, S. W. (2007). The polyvagal perspective. Biological psychology, 74(2), 116-143.
[5] Olmi, B., Frassineti, L., Lanata, A., & Manfredi, C. (2021). Automatic Detection of Epileptic Seizures in Neonatal Intensive Care Units Through EEG, ECG and Video Recordings: A Survey. IEEE Access, 9, 138174-138191.
[6] Mesbah, M., Balakrishnan, M., Colditz, P. B., & Boashash, B. (2012). Automatic seizure detection based on the combination of newborn multi-channel EEG and HRV information. EURASIP Journal on Advances in Signal Processing, 2012, 1-14.
[7] Lyu, J., Shi, W., Zhang, C., & Yeh, C. H. (2023). A Novel Sleep Staging Method based on EEG and ECG Multimodal Features Combination. IEEE Transactions on Neural Systems and Rehabilitation Engineering.