An efficient technique for the development of computer programs for electrocardiogram interpretation.

In this paper is presented an efficient technique for the development of electrocardiogram (ECG) interpretation software, executed rapidly on inexpensive specialized mini or microcomputers; the method--based on the author's experience--is under medical control. The first part of the paper presents the ideal characteristics of an ECG computer interpretation program which in the author's opinion are not yet implemented by any of the present equipment of ECG analysis. The second part presents the author's approach of the problem and the last part gives a description of the programming techniques used.

The logical structure of electrocardiographic diagnosis.

The description of a mathematical model of structure of electrocardiographic (ECG) diagnosis is given, starting from the automatically measurable characteristics of bioelectrical currents generated in the course of cardiac activity. The "dynamic diagnosis", established for rhythm and for atrioventricular conduction disturbances--by analysis in real time of the periodicity of electrical phenomena--is differentiated from the "static diagnosis", established for the other situations--by the logical intersection of amplitude conditions with duration of the waves belonging to a given ECG lead. Likewise, mathematical rationale separates atrial from infraatrial electrical activity.

A programme for computer-aided recognition of electrocardiographic patterns.

A programme for pattern recognition and automatic diagnosis on digitally converted electrocardiograms is described. Pattern recognition is carried out on the 12 classical leads and diagnosis is given in the Minnesota Code, 1968 variant. The programme was written in FORTRAN and tested in a FELIX C 256 computer.