Continuous non-invasive estimates of cerebral blood flow using electrocardiography signals: a feasibility study

Samuel J van Bohemen; Jeffrey M Rogers; Philip C Boughton; Jillian L Clarke; Joaquin T Valderrama; Andre Z Kyme

doi:10.1007/s13534-023-00265-z

Continuous non-invasive estimates of cerebral blood flow using electrocardiography signals: a feasibility study

Biomed Eng Lett. 2023 Feb 9;13(2):185-195. doi: 10.1007/s13534-023-00265-z. eCollection 2023 May.

Authors

Samuel J van Bohemen¹, Jeffrey M Rogers^{2

3}, Philip C Boughton^{4

5}, Jillian L Clarke⁶, Joaquin T Valderrama^{7

8}, Andre Z Kyme^{1

9}

Affiliations

¹ School of Biomedical Engineering, The University of Sydney, Sydney, NSW Australia.
² Neurocare Group, Sydney, NSW Australia.
³ Department of Clinical Medicine, Macquarie University, Sydney, NSW Australia.
⁴ Sydney Spine Institute, Sydney, NSW Australia.
⁵ Sydney Pharmacy School, Faculty of Medicine and Health, The University of Sydney, Sydney, NSW Australia.
⁶ Faculty of Medicine and Health, The University of Sydney, Sydney, NSW Australia.
⁷ National Acoustic Laboratories, Sydney, NSW Australia.
⁸ Linguistic Department, Macquarie University, Sydney, NSW Australia.
⁹ Brain and Mind Centre, The University of Sydney, Sydney, NSW Australia.

Abstract

This paper describes a potential method to detect changes in cerebral blood flow (CBF) using electrocardiography (ECG) signals, measured across scalp electrodes with reference to the same signal across the chest-a metric we term the Electrocardiography Brain Perfusion index (EBPi). We investigated the feasibility of EBPi to monitor CBF changes in response to specific tasks. Twenty healthy volunteers wore a head-mounted device to monitor EBPi and electroencephalography (EEG) during tasks known to alter CBF. Transcranial Doppler (TCD) ultrasound measurements provided ground-truth estimates of CBF. Statistical analyses were applied to EBPi, TCD right middle cerebral artery blood flow velocity (rMCAv) and EEG relative Alpha (rAlpha) data to detect significant task-induced changes and correlations. Breath-holding and aerobic exercise induced highly significant increases in EBPi and TCD rMCAv (p < 0.01). Verbal fluency also increased both measures, however the increase was only significant for EBPi (p < 0.05). Hyperventilation induced a highly significant decrease in TCD rMCAv (p < 0.01) but EBPi was unchanged. Combining all tasks, EBPi exhibited a highly significant, weak positive correlation with TCD rMCAv (r = 0.27, p < 0.01) and the Pearson coefficient between EBPi and rAlpha was r = - 0.09 (p = 0.05). EBPi appears to be responsive to dynamic changes in CBF and, can enable practical, continuous monitoring. CBF is a key parameter of brain health and function but is not easily measured in a practical, continuous, non-invasive fashion. EBPi may have important clinical implications in this context for stroke monitoring and management. Additional studies are required to support this claim.

Supplementary information: The online version contains supplementary material available at 10.1007/s13534-023-00265-z.

Keywords: Cerebral blood flow; Continuous monitoring; Electrocardiography; Electroencephalography; Quantitative electroencephalography; Stroke.