From peripheral finger-derived pulse waveforms to aortic pressure waveform features: an application of a generalized...

Aortic (central) pressure features are associated with cardiovascular complications and can be algorithmically derived from non-invasive peripheral arterial waveforms. This has conventionally been performed with a pressure waveform (i.e., tonometry or oscillometry) rather than with the optical-based sensor (photoplethysmography (PPG)) that is predominantly used in wearable health devices. Extraction of aortic features from a peripheral PPG waveform has yet to be investigated. This study aims to compare aortic features extracted from peripheral arterial waveforms acquired with different sensor modalities using the same transfer function.

Radial tonometry (reference), finger volume-clamped PPG (Peňáz) and fingertip PPG waveforms were measured in participants (n=29, 36±16 years, 15 female) under baseline conditions. Waveforms were converted into an aortic pressure waveform using the transfer function. Waveform features were extracted from the converted waveform. Extracted features were compared with correlation plots and a Bland-Altman analysis.

Aortic pressure features extracted from a finger using the Peňáz technique were comparable to radial tonometry derived features. Aortic features extracted from a fingertip waveform were more variable in comparison to radial tonometry-derived features.

Aortic (central) pressure waveform features contain valuable haemodynamic information and have the capacity to be easily and conveniently implemented in wearable health devices. Future use of these features in wearable health devices incorporating PPG requires the development, and/or, optimization of a unique transfer function to more accurately represent the aortic pressure waveform for cardiovascular assessment.Clinical Relevance- Aortic pressure features might be used in wearable health devices following the development of a unique transfer function for optical-transduced peripheral vascular signals.