AUTHOR=Hellqvist Henrik , Karlsson Mikael , Hoffman Johan , Kahan Thomas , Spaak Jonas 

TITLE=Estimation of aortic stiffness by finger photoplethysmography using enhanced pulse wave analysis and machine learning

JOURNAL=Frontiers in Cardiovascular Medicine

VOLUME=Volume 11 - 2024

YEAR=2024

URL=https://www.frontiersin.org/journals/cardiovascular-medicine/articles/10.3389/fcvm.2024.1350726

DOI=10.3389/fcvm.2024.1350726

ISSN=2297-055X

ABSTRACT=Introduction: Aortic stiffness plays a critical role in the evolution of cardiovascular diseases, but the 20 assessment requires special equipment. Photoplethysmography (PPG) and single-lead ECG are 21 readily available in healthcare and wearable devices. We studied whether a brief PPG registration, 22 alone or combined with single-lead ECG, could be used to reliably estimate aortic stiffness. 23 Methods: A proof-of-concept study with simultaneous high-resolution index finger recordings of 24 infrared PPG, single-lead ECG, and finger blood pressure (Finapres) was performed in 33 subjects 25 (median age 44 [range 21-66] years, 19 men) and repeated within two weeks. Carotid-femoral pulse 26 wave velocity (cfPWV; two-site tonometry with SphygmoCor) was used as reference. A brachial 27 single-cuff oscillometric device assessed aortic pulse wave velocity (aoPWV; Arteriograph) for 28 further comparisons. We extracted 136 established PPG waveform features and engineered 13 new 29 with improved coupling to the finger blood pressure curve. Height-normalized pulse arrival time 30 (NPAT) was derived using ECG. Machine learning was performed to develop prediction models. 31 Results: The best PPG based models predicted cfPWV and aoPWV well (root mean square error 32 0.70 and 0.52 m/s, respectively), with minor improvements by adding NPAT. Repeatability and 33 agreement were on par with the reference equipment. A new PPG feature, an amplitude ratio from 34 the early phase of the waveform, was most important in modelling, showing strong correlations with 35 cfPWV and aoPWV (r=-0.81 and -0.75, respectively, both P<0.001). 36 Conclusion: Using new features and machine learning, a brief finger PPG registration can estimate 37 aortic stiffness without additional information on age, anthropometry, or blood pressure needed. 38 Repeatability and agreement were comparable to non-invasive reference equipment. Provided further 39 validation, this readily available simple method could improve cardiovascular risk evaluation, 40 treatment, and prognosis. 41 42 tog bort: is rarely assessed since it 43 tog bort: We studied whether simple photoplethysmography