Home recording for pre-phase sleep apnea diagnosis by Holter recorder using MMC memory

Long time data recordings by ambulatory devices became common today with widespread use of high capacity memories. In this study, this technology is used on home recording device especially designed for the pre-phase sleep apnea analysis at home before having certain diagnosis by proper polysomnographic examination in a hospital. Designing a portable recording system for clinically significant signals and analyzing an associated apnea detection algorithm are two significant and complementary studies carried out and reported in this paper. According to our aim, the Holter device discussed here is capable of recordings from many channels simultaneously but specifically for this study we consider just three signals namely ECG, breathing activity and oxygen saturation which are all significant for apnea detection in polysomnographic diagnostic phase. In apnea detection part we have initially studied the algorithm based on Hilbert Transform of ECG signal. The study also includes the comparison of three selected QRS detection algorithms (digital filter based, differentiation based and higher order statistics based) for ECG processing required for the Hilbert transform based apnea detection algorithm. Differentiation based QRS algorithm is preferred over other two due to its better performance on noisy ECG signals.