Exploring Arterial Wave Frequency Features for Vascular Age Assessment through Supervised Learning with Risk Factor Insights
With aging being a major non-reversible risk factor for cardiovascular disease, the concept of Vascular Age (VA) emerges as a promising alternate measure to assess an individual’s cardiovascular risk and overall health. This study investigated the use of frequency features and Supervised Learning (S...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-09-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/13/19/10585 |
| _version_ | 1797576286664654848 |
|---|---|
| author | Eugenia Ipar Leandro J. Cymberknop Ricardo L. Armentano |
| author_facet | Eugenia Ipar Leandro J. Cymberknop Ricardo L. Armentano |
| author_sort | Eugenia Ipar |
| collection | DOAJ |
| description | With aging being a major non-reversible risk factor for cardiovascular disease, the concept of Vascular Age (VA) emerges as a promising alternate measure to assess an individual’s cardiovascular risk and overall health. This study investigated the use of frequency features and Supervised Learning (SL) models for estimating a VA Age-Group (VAAG), as a surrogate of Chronological Age (CHA). Frequency features offer an accessible alternative to temporal and amplitude features, reducing reliance on high sampling frequencies and complex algorithms. Simulated subjects from One-dimensional models were employed to train SL algorithms, complemented with healthy in vivo subjects. Validation with real-world subject data was emphasized to ensure model applicability, using well-known risk factors as a form of cardiovascular health analysis and verification. Random Forest (RF) proved to be the best-performing model, achieving an accuracy/AUC score of 66.5%/0.59 for the in vivo test dataset, and 97.5%/0.99 for the in silico one. This research contributed to preventive medicine strategies, supporting early detection and personalized risk assessment for improved cardiovascular health outcomes across diverse populations. |
| first_indexed | 2024-03-10T21:50:05Z |
| format | Article |
| id | doaj.art-ec744f38e75f49499f6f0d25ae6453e9 |
| institution | Directory Open Access Journal |
| issn | 2076-3417 |
| language | English |
| last_indexed | 2024-03-10T21:50:05Z |
| publishDate | 2023-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj.art-ec744f38e75f49499f6f0d25ae6453e92023-11-19T14:01:24ZengMDPI AGApplied Sciences2076-34172023-09-0113191058510.3390/app131910585Exploring Arterial Wave Frequency Features for Vascular Age Assessment through Supervised Learning with Risk Factor InsightsEugenia Ipar0Leandro J. Cymberknop1Ricardo L. Armentano2GIBIO, Facultad Regional Buenos Aires, Universidad Tecnológica Nacional, Buenos Aires C1179AAQ, ArgentinaGIBIO, Facultad Regional Buenos Aires, Universidad Tecnológica Nacional, Buenos Aires C1179AAQ, ArgentinaDepartment of Biological Engineering, Universidad de la República, Paysandú 60000, UruguayWith aging being a major non-reversible risk factor for cardiovascular disease, the concept of Vascular Age (VA) emerges as a promising alternate measure to assess an individual’s cardiovascular risk and overall health. This study investigated the use of frequency features and Supervised Learning (SL) models for estimating a VA Age-Group (VAAG), as a surrogate of Chronological Age (CHA). Frequency features offer an accessible alternative to temporal and amplitude features, reducing reliance on high sampling frequencies and complex algorithms. Simulated subjects from One-dimensional models were employed to train SL algorithms, complemented with healthy in vivo subjects. Validation with real-world subject data was emphasized to ensure model applicability, using well-known risk factors as a form of cardiovascular health analysis and verification. Random Forest (RF) proved to be the best-performing model, achieving an accuracy/AUC score of 66.5%/0.59 for the in vivo test dataset, and 97.5%/0.99 for the in silico one. This research contributed to preventive medicine strategies, supporting early detection and personalized risk assessment for improved cardiovascular health outcomes across diverse populations.https://www.mdpi.com/2076-3417/13/19/10585vascular agemachine learningarterial pressure waveform |
| spellingShingle | Eugenia Ipar Leandro J. Cymberknop Ricardo L. Armentano Exploring Arterial Wave Frequency Features for Vascular Age Assessment through Supervised Learning with Risk Factor Insights Applied Sciences vascular age machine learning arterial pressure waveform |
| title | Exploring Arterial Wave Frequency Features for Vascular Age Assessment through Supervised Learning with Risk Factor Insights |
| title_full | Exploring Arterial Wave Frequency Features for Vascular Age Assessment through Supervised Learning with Risk Factor Insights |
| title_fullStr | Exploring Arterial Wave Frequency Features for Vascular Age Assessment through Supervised Learning with Risk Factor Insights |
| title_full_unstemmed | Exploring Arterial Wave Frequency Features for Vascular Age Assessment through Supervised Learning with Risk Factor Insights |
| title_short | Exploring Arterial Wave Frequency Features for Vascular Age Assessment through Supervised Learning with Risk Factor Insights |
| title_sort | exploring arterial wave frequency features for vascular age assessment through supervised learning with risk factor insights |
| topic | vascular age machine learning arterial pressure waveform |
| url | https://www.mdpi.com/2076-3417/13/19/10585 |
| work_keys_str_mv | AT eugeniaipar exploringarterialwavefrequencyfeaturesforvascularageassessmentthroughsupervisedlearningwithriskfactorinsights AT leandrojcymberknop exploringarterialwavefrequencyfeaturesforvascularageassessmentthroughsupervisedlearningwithriskfactorinsights AT ricardolarmentano exploringarterialwavefrequencyfeaturesforvascularageassessmentthroughsupervisedlearningwithriskfactorinsights |