Fractal dimension analysis for the detection of ventricular late potentials in the time domain

The presence of ventricular late potential (LP) activity in patients with post-myocardial infarction has been correlated with their susceptibility to develop ventricular tachycardia and may be used as a risk criterion. A reliable novel approach for characterising and detecting abnormal LP activity is presented. It consists of analyzing the degree of complexity of LP activity by quantifying their fractal dimension (δ). Fractal dimensional analysis was applied to the LP´s attractor traced in the microvoltage, 3-dimensional space. The δ parameter may be interpreted as an index for the chaotic behaviour of the terminal activity of the ventricular depolarisation process which could trigger a catastrophic arrhythmic episode. Results of a clinical study indicated that the value of δ was significantly higher in at risk patients than in healthy subjects (p<0.001), and suggested that a value of δ>1.30 can be a risk criterion