Academia.eduAcademia.edu

Figure 8 – uploaded by Jungkuk Kim

See full PDFdownloadDownload figure
Fig. 9. Difference between original signal and reconstructed signal of recorc No. 116 at different window sizes. (a) Original data. (b) 80% window. (c) 50% window. (d) 30% window. (e) 15% window.

Figure 9 Difference between original signal and reconstructed signal of recorc No. 116 at different window sizes. (a) Original data. (b) 80% window. (c) 50% window. (d) 30% window. (e) 15% window.

Related Figures (14)

Fig. |. Block diagram of compression and decompression procedures.
Fig. |. Block diagram of compression and decompression procedures.
Fig. 3. Peak-detecting procedure (record no. 111). (a) R-peak detecting of original ECG signal. (b) Peak detection of differenced ECG. (c) Classification of peak interval using difference data. Fig. 2. Flowchart of difference offset calculation procedures for making 1-byte data without loss.
Fig. 3. Peak-detecting procedure (record no. 111). (a) R-peak detecting of original ECG signal. (b) Peak detection of differenced ECG. (c) Classification of peak interval using difference data. Fig. 2. Flowchart of difference offset calculation procedures for making 1-byte data without loss.
Fig. 4. Signals reconstructed with and without error compensation. Very low frequency distortion can be seen in the reconstructed data without error com- pensation. (a) Original data. (b) Without error compensation. (c) With error compensation.
Fig. 4. Signals reconstructed with and without error compensation. Very low frequency distortion can be seen in the reconstructed data without error com- pensation. (a) Original data. (b) Without error compensation. (c) With error compensation.
DATA FORMAT FOR ORIGINAL SIGNAL REPRESENTATION USING SAMPLE DIFFERENCES TABLE I
DATA FORMAT FOR ORIGINAL SIGNAL REPRESENTATION USING SAMPLE DIFFERENCES TABLE I
Fig. 5. Block diagram for error compensating method of cascaded data.
Fig. 5. Block diagram for error compensating method of cascaded data.
Fig. 7. Data-type changing flow in the compression process. Fig. 6. Data trail after calculating cumulative error compensate.
Fig. 7. Data-type changing flow in the compression process. Fig. 6. Data trail after calculating cumulative error compensate.
PERFORMANCE OF PROPOSED COMPRESSION ALGORITHM (20% DCT WINDOW SIZE) TABLE II direction as shown on the right side of Fig. 1. The first proce- dure for reconstruction is to apply the inverse Huffman coding, which refers to the defined data table, as frequently occurs. The second procedure is inverse DCT, and the third procedure is to reconstruct the 1/2 down-sample using the cubic spline algo- rithm. Backward difference reconstruction is the last procedure implemented using (1) and (2), where X,,.(0) is the first restored data instance obtained by adding 2 bytes of the first original data, (start) in Table II and the first differenced data V), (0). The next original sample is obtained using (2), where the next difference data are added to the reconstructed value during the previous step, and so on:
PERFORMANCE OF PROPOSED COMPRESSION ALGORITHM (20% DCT WINDOW SIZE) TABLE II direction as shown on the right side of Fig. 1. The first proce- dure for reconstruction is to apply the inverse Huffman coding, which refers to the defined data table, as frequently occurs. The second procedure is inverse DCT, and the third procedure is to reconstruct the 1/2 down-sample using the cubic spline algo- rithm. Backward difference reconstruction is the last procedure implemented using (1) and (2), where X,,.(0) is the first restored data instance obtained by adding 2 bytes of the first original data, (start) in Table II and the first differenced data V), (0). The next original sample is obtained using (2), where the next difference data are added to the reconstructed value during the previous step, and so on:
Fig. 8. Result of algorithm performance evaluation (average and standard deviation value).
Fig. 8. Result of algorithm performance evaluation (average and standard deviation value).
PERFORMANCE COMPARISON OF DIFFERENT LOSS-TYPE ECG COMPRESSION SCHEMES difference from the original signal when a window size longer than 30% is used.
PERFORMANCE COMPARISON OF DIFFERENT LOSS-TYPE ECG COMPRESSION SCHEMES difference from the original signal when a window size longer than 30% is used.
COMPARISON BETWEENT THE PROPOSED METHOD AND OTHER TYPE COMPRESSION ALGORITHM FOR RECORD NO. 117
COMPARISON BETWEENT THE PROPOSED METHOD AND OTHER TYPE COMPRESSION ALGORITHM FOR RECORD NO. 117
EVALUATION OF STATISTICAL SIGNIFICANCE DEPENDING ON INCORRECT PEAK DETECTION (WINDOW SIZE 50%, FOR RECORD No. 100) TABLE VI
EVALUATION OF STATISTICAL SIGNIFICANCE DEPENDING ON INCORRECT PEAK DETECTION (WINDOW SIZE 50%, FOR RECORD No. 100) TABLE VI
Fig. 10. Example for e-health data transmission system using proposed com- pression algorithm. EVALUATION OF STATISTICAL SIGNIFICANCE DEPENDING ON INCORRECT PEA} DETECTION (WINDOW SIZE 20%, FOR RECORD No. 100)
Fig. 10. Example for e-health data transmission system using proposed com- pression algorithm. EVALUATION OF STATISTICAL SIGNIFICANCE DEPENDING ON INCORRECT PEA} DETECTION (WINDOW SIZE 20%, FOR RECORD No. 100)
Related topics:
AlgorithmsGeriatricsBiomedical EngineeringSignal ProcessingData CompressionComputer SecurityBiomedical Signal ProcessingFrequency-domain analysisReal Time SystemsData Encryption
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved