This paper will present a proposed hardware design of an intelligent device for ECG signal acquisition and classification. The device will be able to collect one channel of ECG signal, analyze it and classifY the signal to detect the arrhythmias using a neuro-fuzzy TSK network. The classification model was trained by using the sample ECG signals taken from the MIT-BIH database (available on physionet.org). The device's design will use the programmable IC technologies such as FPAA (Field Programmable Analog Array) and PSoC (Programmable System on Chip) in order to: 1/ tune the parameters of the circuits in a easy way for better performance
2/ implement the classification model, which is the neuro-fuzzy TSK (Takagi Sugeno - Kang) network, especially in the case when the authors need to update the structure and parameters of network due to new results from a learning process to adapt the network to a new set of data. Also thanks to the high degree of functional blocks integration on the FPAA and PSoC ICs, the authors can create portable devices with very compact size and easy to use. A very strong advantage of FPAA is the fact that the parameters of the filters and the amplifiers implemented in FPAA can be changed "in fly" during the working process of the circuit. This capability allows us to create adaptive, flexible devices for different working conditions and environments.