Classification of Hyperkinetic, Hypokinetic and Normal Segments in Ventriculographic Images, using Centerline Method

Abstract

Evaluation of regional left ventricular wall motion is of vital importance at the clinical level, since this cavity is the most susceptible to severe damage in diseases such as arterial hypertension, diabetes mellitus and atherosclerosis. This paper reports on the design of a software platform for the estimation of parameters describing the left ventricular wall motion in ventriculographic images. The system focuses on end-diastolic and end-systolic contours. Firstly is carried out a process of manual segmentation by the specialist, then 100 chords generated between two ventricular contours are quantified, which allows the classification of these segments (normal, hyperkinetic and hypokinetic). At last, the testing process is performed using actual data acquired in the Institute Autonomy Hospital Universitary of University of the Andes IAHULA.

Keywords

• pared ventricular izquierda
• sistema semiautomático
• imágenes ventriculográficas
• hyperkinésis
• hypokinésis