| Description |
The platform Calicot (Cardiac Arrhythmias Learning for Intelligent Classification of On-line Tracks) integrates the different software components used for learning and recognition of cardiac arrhythmias.
The signal processing modules implemented in Matlab provide detection and classification of waves from an ECG to produce a stream of symbolic time-stamped events. This flow can be placed at the input of the chronicle recognizer CRS in order to perform the detection of cardiac arrhythmias. CRS, made at our disposal by France Telecom R & D (http://crs.elibel.tm.fr/), comes in the form of a library of C++ programs from which it is possible to write chronicle recognizers adapted to various application domains. The stream of symbolic events can also be used to build an example database from chronicles will be learned. Learning is achieved by means of ICL (http://www.cs.kuleuven.ac.be/~wimv/ICL/), a machine learning system using inductive logic programming. |
