The implantable cardioverter-defibrillator (ICD) is an effective therapeutic device for rescuing patients with cardiac diseases from death caused by life-threatening arrhythmias. The authors previously have proposed a detection algorithm of life-threatening arrhythmia with a multiple regression model. In this research, we have developed the algorithm so as to accurately classify cardiac rhythms and to reduce in the computational time with a microcontroller used in the ICD. The experimental results showed that the proposed method kept a high accuracy to detect cardiac rhythms. In addition, the validation of implementation of the proposed algorithm in the microcontroller indicated that the result of detection of cardiac rhythms could be attained within computational time of 60 ms. For the practical application, it is necessary to evaluate the power consumption of the ICD working with the proposed method.
