These tools will no longer be maintained as of December 31, 2024. Archived website can be found here. PubMed4Hh GitHub repository can be found here. Contact NLM Customer Service if you have questions.


Pubmed for Handhelds

PUBMED FOR HANDHELDS

Search MEDLINE/PubMed

  • Title: Adaptive compression of the ambulatory electrocardiogram.
    Author: Hamilton PS.
    Journal: Biomed Instrum Technol; 1993; 27(1):56-63. PubMed ID: 8418967.
    Abstract:
    Previous use of the MIT/BIH arrhythmia database, on analog tape, to investigate compression of ambulatory ECG data by average beat subtraction, residual differencing, and Huffman coding of the residuals had shown that with a quantization level of 35 mu V and a sample rate of 100 samples per second, it was possible to store ECG data with average data rates of 174 bits per second (bps), but because of the variation in ECG signals, data rates for different records ranged from 144 bps to 230 bps. In a practical storage system, it is desirable to fix the maximum data rate and store data with a minimum of distortion. For this study the previous compression algorithm was modified to adapt its quantization level to different ECG signal conditions. Two adaptation strategies were investigated. Both adapt the quantization-step size according to the number of bytes required for storing the coded signal, beat arrival times, and beat classifications. The new compression algorithm was tested with data from the MIT/BIH database on CD ROM. With the more successful of the two strategies, the adaptive compression algorithm stored MIT/BIH records with a difference of only 0.8 bps between the record with the highest data rate and the record with the lowest data rate. The average data rate for the entire database was 193.3 bps. Signal-to-compression noise ratios varied from record to record and varied over time for a given record. Average signal to compression noise ratios varied from 26.82 to 532.83.
    [Abstract] [Full Text] [Related] [New Search]