臺灣博碩士論文加值系統

心電圖是記錄心臟收縮和擴張運動所產生的微小電氣變化，對於心臟疾病的診斷是一個不可或缺的輔助，醫生可依波形的變化判斷可能罹患的心臟疾病，但是因為儲存12導程心電圖資料需要龐大的空間，具保持臨床診斷資訊能力之心電圖壓縮技術乃因應而生。
本論文使用FPGA實現一目前壓縮效能最高且可即時運算的新式心電圖壓縮系統[1]中的量化演算法。此系統包含3個部分，第一個部分是可逆式四捨五入非遞迴小波轉換，第二個部分是非線性量化演算法，第三個部分是SPIHT編碼[2]。非線性量化演算法的特色在於只須使用一個量化參數QF就能夠得到線性的失真度-壓縮率曲線，而其內部的新式QOD(Quality On Demand)機制則可在頻域中直接進行失真度預測，並保證壓縮品質。本篇論文修改原始非線性量化演算法的部分設計以方便硬體的實現，並以簡單的硬體設計完成非線性量化演算法中的小數除法及小數根號的運算。
本論文的硬體設計可以實現高效率的運算，量化每段1024筆心電圖訊號在硬體QOD機制的遞迴運算下，最多於5次內必定達到使用者設定的心電圖重建品質。而心電圖單一導程的量化處理時間每段最多約為1.092m秒。在考量壓縮12導程之需求下，其量化處理最多約0.013秒。對於1KHz的心電圖取樣訊號而言，系統可有效地達到心電圖的及時記錄。本論文所使用的FPGA晶片為ALTERA Cyclone Ⅱ-EP2C35F672C6N，工作頻率為50MHz，使用的邏輯單元為2383個。

Electrocardiogram (ECG) recording of the electrical activity of the heart is widely used for clinical diagnosis of heart disease. However, storing the 12-lead ECG signal needs huge capacity. In order to keep the ECG signal clinical features, then a lot of ECG
compression methods were proposed.
This study implements the novel compression system [1] on FPGA. The novel compression system has not only the highest compression ratio but also real-time operation at present. This system includes 3 parts: first part is a reversible round-off non-recursive discrete period wavelet transform (RRO-NRDPWT); second part is a non-linear quantization algorithm; third part is a SPIHT algorithm [2]. The non-linear quantization algorithm acquired a linear relation curve between distortion and compression ratio by the single quantization factor (QF).The Quality on Demand (QOD) predicts the distortion directly in frequency domain, and it will guarantee the reconstructed ECG signals for each segment with the desired quality. To implement hardware, this study modifies the original non-linear quantization algorithm. The decimal division and decimal square root in non-linear quantization algorithm were
implemented by simple design.
The hardware of this study has high efficiency. The user can set up the reconstruction quality less than 5 times under the novel QOD method for each segment ECG signals. For the non-linear quantization algorithm, the maximum execution time on FPGA is 1.092ms for single channel ECG signal, and 0.013 seconds for the 12-lead ECG signals. This system records real-time ECG signals effectively for 1KHz sampling rate. Implementing the non-linear quantization algorithm needs 2383 LEs and the
operation frequency is 50MHz on FPGA Cyclone Ⅱ-EP2C35F672C6N.

中文摘要----------------------------------------------- Ⅰ
英文摘要----------------------------------------------- Ⅱ
致謝--------------------------------------------------- Ⅲ
目錄--------------------------------------------------- Ⅳ
表目錄------------------------------------------------- Ⅵ
圖目錄------------------------------------------------- Ⅶ
第一章 緒論------------------------------------------ 1
1.1 研究動機與目的------------------------------------- 1
1.2 心電圖資料壓縮方法回顧----------------------------- 4
1.3 實驗設備------------------------------------------- 5
1.4 論文架構------------------------------------------- 7
第二章 小波轉換理論------------------------------------ 8
2.1 小波轉換------------------------------------------- 8
2.1.1 遞迴式離散週期性小波轉換(DPWT)------------------- 8
2.1.2 非遞迴離散週期性小波轉換(NRDPWT)----------------- 11
2.2 可逆式四捨五入線性轉換(RROLT)---------------------- 14
第三章 心電圖壓縮系統---------------------------------- 16
3.1 基於RRO-NRDPWT的心電圖壓縮系統--------------------- 16
3.2 非線性量化演算法----------------------------------- 18
3.3 壓縮系統失真分析----------------------------------- 22
3.4 壓縮品質控制機制(QOD)------------------------------ 26
3.5 QOD機制之失真度預測-------------------------------- 29
第四章 心電圖壓縮系統量化演算法硬體設計---------------- 31
4.1 量化演算法硬體規劃--------------------------------- 31
4.1.1 量化尺度使用方式--------------------------------- 31
4.1.2 量化尺度有限位元分析----------------------------- 33
4.1.3 量化演算法有限狀態機設計------------------------- 37
4.2 硬體架構及實現方式--------------------------------- 39
4.2.1 讀取小波分解係數--------------------------------- 39
4.2.2 量化尺度讀取------------------------------------- 40
4.2.3 量化處理----------------------------------------- 42
4.2.4 失真度計算--------------------------------------- 44
4.2.5 Step預測計算------------------------------------- 48
第五章 量化演算法電路模擬結果及分析-------------------- 49
5.1 量化處理------------------------------------------- 50
5.2 失真度計算----------------------------------------- 52
5.3 量化階級預測計算----------------------------------- 54
5.4 量化尺度獲得--------------------------------------- 56
5.5 模擬結果分析--------------------------------------- 58
第六章 結論-------------------------------------------- 60
參考文獻----------------------------------------------- 61

[1] Cheng-Tung Ku，A High Efficient ECG Data Compression System Based On
Reversible Round-Off Discrete Wavelet Transform，National Kaohsiung First University of Science and Technology，doctoral dissertation, 2007.


[2] A.Said and W.A. Pearlman, “A new fast, and efficient image codec based on
set partitioning in hierarchical trees”, IEEE Transactions on Circuits and System for Video Technology,vol.6,No.3,pp.243-250,Jun. 1996.


[3] B. R. Shankara and I. S. N. Murthy, “ECG data compression using Fourier
descriptors,” IEEE Transaction on Biomedical Engineering, vol. 33, pp. 428–433, Apr. 1986.


[4] T. Blanchett, G. C. Kember, G. A. Fenton, “KLT-based quality controlled
compression of single-lead ECG,” IEEE Transactions on Biomedical Engineering, vol. 45, no. 7, pp. 942-945 July. 1998.


[5] L. Batista, E. Melcher, and L. Carvalho, “Compression of ECG signals by
optimized quantization of discrete cosine transform coefficients,” Medical Engineering & Physics, vol. 23, no.2, pp. 127-134, March. 2001.


[6] R. Benzid, F. Marir, A. Boussaad, M. Benyoucef , and D. Arar, “Fixed percentage
of wavelet coefficients to be zeroed for ECG compression,” Electronics Letters, vol. 39, no. 11, pp. 830-831, May 2003.



[7] J. M. Shapiro, “Embedded image coding using zerotrees of wavelet coefficients,”
IEEE Transactions on Signal Processing, SP-41, pp. 3445-3462, December 1993.

[8] B.-U. Kohler, C. Hennig, R. Orglmeister, “The principles of software QRS
detection,” IEEE Engineering in Medicine and Biology Magazine, vol. 21, no. 1, pp. 42-57, Jan./Feb. 2002.


[9] Chia-Chun Sun, Shen-Chuan Tai,”Beat-Based ECG Compression Using
Gain-Shape Vector Quantization”, Biomedical Engineering, IEEE Transactions on Volume 52, Issue 11, Nov. 2005 Page(s):1882 – 1888


[10] I. Daubechies, “Ten lectures on wavelet,” Ser. No.61 in CBMS-NSF series in
Applied Mathematics. Philadelphia, PA: SIAM, 1992.


[11] C.-T. Ku, H.-S. Wang, K.-C. Hung, and Y.-S. Hung. “A novel ECG data
compression method based on nonrecursive discrete periodized wavelet transform,” IEEE Transactions on Biomedical Engineering, vol. 53, no. 12, pp. 2577-2583, Dec. 2006.


[12] S.-G. Miaou, and C.-L. Lin, “A quality-on-demand algorithm for wavelet-based
compression of electrocardiogram signals,” IEEE Transactions on Biomedical Engineering, vol. 49, no. 3, pp. 233-239, March 2002.

[13] 麻省理工學院心律不整心電圖資料庫:
http://www.physionet.org/physiobank/database/mitdb/