چكيده به لاتين
This research has been dedicated the growing need of the human community for early detection of heart disease. Heart disease remains the first cause of human death for many years. In this study, we looked into more accurate and faster solutions for the automatic diagnosis of cardiac arrhythmias. To this end, the following procedure has been considered:
First, we introduced a reliable algorithm for the automatic detection of cardiac arrhythmias from the ECG. The algorithm consists of three main steps: pre-processing, feature extraction, and finally the classification of the ECG signal into several types of arrhythmias. We verified the algorithm's performance with test on the beats in the MIT-BIH database. To understand the concepts of the selected algorithm, we reproduced the steps of the algorithm while improving procedures like detection and classification of the heartbeat signal. The optimized algorithm is obtained with less complexity and more accuracy than the base algorithm.
Then, to achieve the goal of this study, we implemented an automatic detection algorithm in hardware. To do this, we first designed and built a device with the ability to record and store the ECG, so that we could observe the performance of the algorithm objectively. After the recording of the ECG, the signal is sent to the computer and the algorithm is applied. We also designed and built a device that performs all the steps of recording, preprocessing, feature extraction and Arrythmia Classification. The stages of detection are performed within two microcontrollers of the ARM family.
Finally, we presented a method for the automatic detection of arrhythmia remotely in the cloud. In this method the recorded ECG will be transferred to a server for automated detection, and diagnostic operation will be performed in the server. This is done by use of a smartphone. By performing this process, the automatic detection of arrhythmia will no longer be limited to the presence of the doctors and will speed the diagnosis process.
Keywords:
ECG, Arrhythmia, Automatic detection, hardware implementation, cloud base detection
