Revista argentina de cardiología
versión On-line ISSN 1850-3748
INTROZZI, Aníbal R. et al. Relation between Aging and the Morphology of Radial Artery Diameter Variation Registers: Classification Using Artificial Neural Networks. Rev. argent. cardiol. [online]. 2005, vol.73, n.5, pp. 352-357. ISSN 1850-3748.
The aim of this study was to assess the modification of the radial artery diameter variation signal in accordance with aging. The signal was registered via an acquisition system using a capacitive transducer leaning on the skin at the level of the radial artery. In order to assess the relation between loss of distensibility and signal modifications due to aging, the experiment was carried out on a group of 64 normotensive subjects: 20 youths, 24 adults and 20 elderly men. Morphological descriptive parameters were obtained and the potentiality to classify the arterial diameter variation waves according to aging was evaluated with a Kohonen artificial neural network. Signal analysis processing revealed the presence of three constituent waves: two systolic waves, first and second, and a diastolic wave. Aging produces important changes in signal conformation, such as expansion of the first systolic wave, decrease of the diastolic wave amplitude and rise of the second systolic wave approaching and overlapping the vertex of the first systolic wave. The artificial neural network correctly classified the 64 registrations, locating them in their corresponding age cluster. Considering the absence of illness in the sample subjects, the existence of a healthy arterial aging pattern is revealed. This system could be able to detect early evidence of arterial pathology development involving an increment of arterial stiffness with a loss of compliance (sclerosis, hypertension).
Palabras clave : Artery; Diameter; Age; Neural Networks (Computer).