Automatic monitoring system for artificial hearts using self organizing map.

This study presents an automatic monitoring system for artificial hearts. The self organizing map (SOM) was applied to monitoring and analysis of an aortic pressure (AoP) signal measured from an adult goat equipped with a total artificial heart. In the proposed system, two different SOMs were used to detect and classify abnormalities in the measured AoP signal. In the first stage, an ordinary SOM, taught with only normal AoP data, was used for detection of abnormalities on the basis of the quantization error in the real-time monitoring task. In the second stage, a supervised SOM was used for classification of abnormalities. The supervised SOM can be regarded as an ordinary SOM with an extra class vector for solving the classification problem. The class vector is assigned to every node in the second SOM as an output weight learned according to Kohonen's learning rule. The effectiveness of detection and classification of abnormalities using these two SOMs was confirmed.