To study the atherosclerotic lesion (a.l.) in the two aorta segments (thoracic and abdominal) in a given set of autopsies (total 2043) performed during five years (1981-85) in one of the principal general hospitals in the city of Havana, a set of five variables was used as part of a new biometric system (BS), to characterize the a.l. in any vascular sector. Three of these variables represent the main types of a.l., fatty streaks (X), fibrous plaques (Y) and severe (complicated and calcified) plaques (Z), while the others were indices of stenosis (P) and benignity (B). Classical dissection and pathological laboratory procedures were performed. Qualitative and quantitative gross morphometric analyses were done by a digitilizer joined to a NEC (9801) personal microcomputer (Japan). These data were processed in a medium size computer EC-1040 (GDR). The multivariate statistical techniques, the principal component analysis (PCA) and the discriminant analysis (DA), were used applying the "SPSS" programme.
Conclusions: (1) PCA revealed in the two aorta segments studied a first component of benignity, fatty streaks (X) and benignity index (B) and a second component of severity, severe plaques (Z) and stenosis number (P). Because of the dimensions of the two aorta segments (width and length) the dominant, first component, is benignity. So the BS is useful in characterizing and describing the lesional state of any aorta segments; (2) DA and BS proved to be useful to distinguish between the high atherosclerotic group (HAG) and the low atherosclerotic group (LAG). The correct classification rate exceed in all cases 70%. The three components of the lesional state vector (X,Y,Z) distinguish the groups. The stenosis (P) and benignity (B) indices proved to be most relevant. The sign of benignity index is always the same as that of the LAG. Thus this variable truly represents benignity. The coherence and consistency of the BS was also proved by DA.