Development of an electrocardiographic analysis algorithm
Thesis DisciplineElectrical Engineering
Degree GrantorUniversity of Canterbury
Degree NameMaster of Engineering
There is a need to automate the present manual methods of analysing long term (24 hour) recorded Electrocardiogram (ECG) signals at high speed (60 times the recorded speed). An algorithm has been developed which groups the recorded heart beats into one of up to six families depending on its shape and also into 'early/on time/late' depending on the time of occurrence of a beat with respect to the precedding beats. The algorithm was implemented on a general purpose mini-computer and supplied with data from the ECG tape replay unit. A total of 4608 beats so analysed were manually checked and found to be correct. The construction of a stand-alone microprocessor implementation of the algorithm is proposed and possible extensions to the algorithm are discussed.