Research Progress of MRI Gradient Waveform Generator

doi:10.11938/cjmr20233067

Abstract

Abstract:

The accurate encoding of the spatial position of imaging voxels is crucial for the spatial localization of magnetic resonance imaging (MRI). To guarantee spatial encoding accuracy, it is essential to obtain high-performance gradient magnetic fields. Gradient waveform generators are the core part of generating gradient magnetic fields. This paper briefly introduces the working principle and components of gradient waveform generator, summarizes the research progress of gradient waveform generator based on different design schemes in the past 20 years, and discusses the realization methods in detail of gradient waveform pre-emphasis. Besides, the future research and development direction of gradient waveform generators is prospected.

Key words: magnetic resonance imaging (MRI), gradient waveform generator, DSP, FPGA, pre-emphasis

CLC Number:

O482.53

LIU Ying, LIN Ling, YUAN Binhua, ZHANG Haowei. Research Progress of MRI Gradient Waveform Generator[J]. Chinese Journal of Magnetic Resonance, 2024, 41(1): 99-115.

Figures/Tables 12

Fig. 1

Fig. 2

Fig. 3

Table 1

Fig. 4

Fig. 5

Fig. 6

Fig. 7

Table 2

Comprehensive comparison of various design schemes

设计方案	第一作者	关键技术	主时钟频率	DAC位宽	存储器容量	精度、时间分辨率及场强	优势和劣势
基于DSP的设计方案	Dai^[7]	DSP（TMS320C6713B，TI）基于PCI总线，接收来自上位机的数据并预处理，DSP 代码编译环境为 CCS 2.0	300 MHz	PCM1704，采样频率16~96 kHz， 24 bit	外接SDRAM 与FLASH： 32 MB/8 MB	数据精度32 bit， 0.3 T	性能稳定、成本低廉；灵活性、可扩展性不足，实时性较差^[24]
基于FPGA的设计方案	Kumar^[5]	采用FPGA构建处理器及梯度波形合成器，使用Vivado软件进行设计与仿真	/	未说明所选用的DAC	FIFO IP核，未说明数据深度	输出梯度波形的最小分辨率为0.4 ms	节省了FPGA 资源，减少了梯度波形生成时间；系统输出速率快，可能导致输出为空
基于FPGA的设计方案	Xing^[3]	使用FPGA （EP2C35F484， Altara）及Quartus II 软件共同设计梯度波形发生器	50 MHz	未说明所选用的DAC	RAM： 483840 bit	数据精度24 bit，时间分辨率1 μs	使用串行运算减少了FPGA 乘法器资源消耗；增加了运算的难度
基于DSP和FPGA结合的设计方案	Tang^[20]	DSP（TMS320LF2407A， TI）将数据等传至 FPGA（EP3C55F484， Altera），与Simulink 和System Generator 软件联合设计梯度波形发生器	/	PCM1704	RAM: 292 KB	数据精度32 bit，时间分辨率1 μs， <0.7 T	电路板尺寸小，成本低；采用 USB进行上位机与梯度波形发生器的通信，通信速度相较以太网而言较慢
	Ai^[33]	上位机由以太网传递信号至FPGA （EP3C40F484C6， Altera），DSP （ADSP21369，ADI）实现梯度计算，软件Quartus II及 Moldelsim	50 MHz	PCM1704	SDRAM： 1GB/4GB	幅度参数精度24 bit，时间参数精度为32 bit，时间分辨率1 μs， <0.5 T	切换精度小于100 μs；采用DSP进行梯度计算在运算速度及计算量上存在不足
	Xiao^[2]	DSP（TMS320VC33，TI）接收数据， FPGA（EP2C8Q208， Altera）实现梯度计算	60 MHz	未说明所选用的DAC	ROM： 256 KB， SDRAM： 1 MB	时间分辨率1 μs， 0.35 T	DSP有32条控制线及24条地址线，有利于对FPGA进行控制及传输信号

Table 2

Table 3

Fig. 8

Table 4

References 66

[1]	WANG X N, LI C M, CHEN M. Opportunities and challenges in magnetic resonance imaging: Development achievements and prospects in China in the past decade[J]. Chin J Magn Reson Imaging, 2022, 13(10): 1-4.
	王笑男, 李春媚, 陈敏. 磁共振成像机遇和挑战——中国十年来发展成果及展望[J]. 磁共振成像, 2022, 13(10): 1-4.
[2]	XIAO L, LI L, NIE W, et al. A digital signal processor-based pulse programmer with performance of run-time information handling for magnetic resonance imaging[J]. Meas Sci Technol, 2015, 26(6): 065402. doi: 10.1088/0957-0233/26/6/065402
[3]	XING C J, WANG T Q, XIAO L. A compact and high-precision gradient waveform generator using an FPGA device for magnetic resonance imaging[C]// 15th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI), Beijing, China: IEEE, 2022: 1-6.
[4]	赵喜平. 磁共振成像系统的原理及其应用[M]. 科学出版社, 2000.
[5]	Khandagle K, Rathod S S. Implementation of MRI gradient generation system and controller on field programmable gate array (FPGA)[C]// International Conference on Communication information and Computing Technology (ICCICT), Mumbai, India: IEEE, 2018: 1-4.
[6]	PAN W Y, ZHANG F, LUO H, et al. Design of MRI gradient calculation module based on high performance DSP[J]. Chin J of Med Instrum, 2011, 35(03):189-193.
	潘文宇, 张富, 罗海, 等. 一种基于高性能DSP的MRI梯度计算模块设计[J]. 中国医疗器械杂志, 2011, 35(3): 189-193.
[7]	DAI Y D, NING R P, LI G Y. Gradient waveform generator based on DSP[J]. Chinese J Magn Reson, 2009, 26(1):44-50.
	戴祎栋, 宁瑞鹏, 李鲠颖. 基于DSP技术的梯度波形发生器[J]. 波谱学杂志, 2009, 26(1): 44-50.
[8]	WANG H W, LI Z G, MAO W P, et al. Design of gradient-field-frequency interlocking-uniform field system for integrated NMR spectrometer[J]. Chinese J Magn Reson, 2012, 29(4):489-498.
	王恢旺, 李正刚, 毛文平, 等. 一体化NMR波谱仪梯度-场频联锁-匀场系统设计[J]. 波谱学杂志, 2012, 29(4): 489-498.
[9]	YAN X L, XIE H B, BIAN M H, et al. A general purpose nuclear magnetic resonance micro-imaging system[J]. Chinese J Magn Reson, 2000, 17(6): 433-439.
	晏小龙, 谢海滨, 边明华, 等. 一种通用的核磁共振微成像系统[J]. 波谱学杂志, 2000, 17(6): 433-439.
[10]	XU Q, WANG H, JIANG Y, et al. A gradient waveform generator for magnetic resonance imaging with digital pre-emphasis[J]. Chinese J Magn Reson, 2006, 23(1): 11-16.
	徐勤, 王鹤, 蒋瑜, 等. 一种具有数字预加重的磁共振成像梯度波形发生器[J]. 波谱学杂志, 2006, 23(1): 11-16.
[11]	OMAR D, AHMED G, HAMID O, et al. Comparison study of hardware architectures performance between FPGA and DSP processors for implementing digital signal processing algorithms: Application of FIR digital filter[J]. Results Eng, 2022, 16: 100639. doi: 10.1016/j.rineng.2022.100639
[12]	戴祎栋. 基于DSP技术的成像谱仪硬件单元的研制[D]. 华东师范大学, 2009.
[13]	HOU S L, XIE H T, HOU X W, et al. Gradient coil and image quality of permanent magnet micro magnetic resonance imaging instrument[J]. Chinese J Magn Reson, 2012, 29(4): 508-520.
	侯淑莲, 谢寰彤, 侯晓吻, 等. 永磁微型磁共振成像仪的梯度线圈与图像质量[J]. 波谱学杂志, 2012, 29(4): 508-520.
[14]	寇松方. 基于PXIe架构的一种新型数字化磁共振成像谱仪梯度控制器的研制[D]. 电子科技大学, 2018.
[15]	CHEN Z H, DENG S Z, XU N S. A scheme of video signal processing system foe field emission display based on DSP and FPGA devices[J]. Acta Scientiarum Naturalium Universitatis Sunyatseni, 2010, 49(2):8-12.
	陈振华, 邓少芝, 许宁生. 一种基于DSP与FPGA实现场发射平板显示器视频信号处理系统的方案[J]. 中山大学学报(自然科学版), 2010, 49(2): 8-12.
[16]	LE H, KHOI N, DINH D, et al. A proposed RISC instruction set architecture for the MAC unit of 32-bit VLIW DSP processor core[C]// International Conference on Computing, Management and Telecommunications (ComManTel), Da Nang, Vietam: IEEE, 2014: 170-175.
[17]	MOHAMMED B, CHRISTOPHE G, JEAN C, et al. Survey on hardware implementation of random number generators on FPGA: Theory and experimental analyses[J]. Comput Sci Rev, 2018, 27: 135-153. doi: 10.1016/j.cosrev.2018.01.002
[18]	RAGIB A M, ABBAS N H. Comparison between FPGA Co-Processor&Tms320 C641x DSP Family in Implementing DIF FFT Algorithm[J]. Journal of Engineering, 2005, 11(3): 533-540.
[19]	SEPULVEDA C, MUNOZ J, ESPINOZA J, et al. FPGA v/s DSP performance comparison for a VSC-based STATCOM control application[J]. IEEE T Ind Inform, 2013, 9(3): 1351-1360. doi: 10.1109/TII.2012.2222419
[20]	TANG W, WANG W. A single-board NMR spectrometer based on a software defined radio architecture[J]. Meas Science& Technol, 2010, 22(1): 015902.
[21]	ELI F. Note:High resolution ultra fast high-power pulse generator for inductive load using digital signal processor[J]. Rev Sci Instrum, 2014, 85(8): 086108. doi: 10.1063/1.4893350
[22]	TAN P, LIU C Y, ZHANG Y, et al. Application of DSP technology in nuclear magnetic resonance data system[J]. Chinese J Magn Reson, 2000, 17(2): 166-172.
	谭萍, 刘朝阳, 张岩, 等. DSP技术在核磁共振数据系统中的应用[J]. 波谱学杂志, 2000, 17(2): 166-172.
[23]	BRYDEN N, Antonacci M, Kelley M, et al. An open-source, low-cost NMR spectrometer operating in the mT field regime[J]. J Magn Reson, 2021, 332: 107076. doi: 10.1016/j.jmr.2021.107076
[24]	XIAO L, TANG W N, WANG W M. Magnetic resonance imaging gradient calculation module based on single chip FPGA[J]. Chinese J Magn Reson, 2010, 27(2): 163-171.
	肖亮, 汤伟男, 王为民. 基于单片FPGA的磁共振成像梯度计算模块[J]. 波谱学杂志, 2010, 27(2): 163-171.
[25]	ZHAO Z, ZHANG H X, LI X Y, et al. Design of optical fiber data transmission circuit based on FPGA+DSP[J]. Modern Electronics Technique, 2023, 46(6):16-22.
	赵喆, 张会新, 李晓宇, 等. 基于FPGA+DSP的光纤数据传输电路设计[J]. 现代电子技术, 2023, 46(6): 16-22.
[26]	ZHANG Z, MAO W P, LIU C Y. Research progress in nuclear magnetic resonance pulse sequence generator[J]. Chinese J Magn Reson, 2012, 29(3): 339-345.
	张志, 毛文平, 刘朝阳. 核磁共振脉冲序列发生器研究进展[J]. 波谱学杂志, 2012, 29(3): 339-345.
[27]	ARIANDO D, CHENG C, GREER M, et al. An autonomous, highly portable NMR spectrometer based on a low-cost System-on-Chip (SoC)[J]. J Magn Reson, 2019, 299: 74-92. doi: S1090-7807(18)30330-6 pmid: 30590351
[28]	俎栋林. 核磁共振成像学[M]. 高等教育出版社, 2004.
[29]	Layton K J, Kroboth S, Jia F, et al. Pulseq: A rapid and hardware-independent pulse sequence prototyping framework[J]. Magn Reson Med, 2017, 77: 1544-1552. doi: 10.1002/mrm.26235 pmid: 27271292
[30]	HUANG C H, ZHANG Z, CHEN L, et al. Design of time division multiplexing for MRI gradient pre-emphasis module[J]. Chinese J Magn Reson, 2018, 35(4): 465-474.
	黄朝晖, 张志, 陈黎, 等. MRI梯度预加重模块的分时复用设计[J]. 波谱学杂志, 2018, 35(4): 465-474.
[31]	HU H C, WANG M S, DAI J S. Design and implementation of dual core serial communication system based on DSP and FPGA[J]. Foreign Electronic Meas Technol, 2022, 41(8): 125-132.
	胡汉昌, 王茂森, 戴劲松. DSP+FPGA的双核串行通信系统设计与实现[J]. 国外电子测量技术, 2022, 41(8): 125-132.
[32]	LIU C, PENG S, LIN X F. Receiver design based on FPGA and DSP architecture[J]. Electronic Design Engineering, 2023, 31(3): 141-144.
	刘成, 彭思, 林选锋. 基于FPGA与DSP架构的接收机设计[J]. 电子设计工程, 2023, 31(3): 141-144.
[33]	艾吉斯. 开放式永磁型MRI谱仪硬件系统的研究与设计[D]. 电子科技大学, 2012.
[34]	YAN F G, LI J T, ZHOU X Y. DSP6711 Ethernet interface extension based on DM9000A[J]. Electronics World, 2019, 578(20):27-28.
	闫法钢, 李江涛, 周晓燕. 基于DM9000A的DSP6711以太网接口扩展[J]. 电子世界, 2019, 578(20): 27-28.
[35]	JOSÉ P M, FRANK F J S, ANDREW G W. Low-field MRI: An MR physics perspective[J]. J Magn Reson Imaging, 2019, 49(6): 1528-1542. doi: 10.1002/jmri.26637 pmid: 30637943
[36]	权怡心. 基于FPGA的嵌入式DSP核的硬件设计与实现[D]. 西安电子科技大学, 2020.
[37]	ZHANG B, LU P, WANG J, et al. A high-speed and area-efficiency DSP block embedded in FPGAs[C] // 13th IEEE International Conference on Solid-State and Integrated Circuit Technology (ICSICT), Hangzhou, China: IEEE, 2016: 1497-1499.
[38]	CEVA and FLEX LOGIX launch DSP chips with embedded FPGAs[J]. Microcontrollers&Embedded Systems, 2022, 22(9): 96.
	CEVA和FLEX LOGIX推出具有嵌入式FPGA的DSP芯片[J]. 单片机与嵌入式系统应用, 2022, 22(9): 96.
[39]	LIU Y, SONG M H, WANG K, et al. An MRI signal receiving method based on radio frequency direct band-pass sampling with system generator[J]. Chinese J Magn Reson, 2018, 35(4): 475-485.
	刘颖, 宋明辉, 王坤, 等. 基于全可编程SoC和LabVIEW的磁共振接收系统设计[J]. 波谱学杂志, 2018, 35(4): 475-485.
[40]	JIANG Z, YANG K C, FISHER N T, et al. AXI-IC^RTRT: Towards a real-time AXI-interconnect for highly integrated SoCs[J]. IEEE T Comput, 2023, 72(3): 786-799.
[41]	LIAO W S, XU J C, YAO S Q, et al. Phase coherence technology of digital MR console based on dual reference sources[J]. Chinese J Magn Reson, 2022, 39(3): 327-336.
	廖文姗, 徐俊成, 姚守权, 等. 基于双参考源的数字磁共振控制台相位相干技术[J]. 波谱学杂志, 2022, 39(3): 327-336.
[42]	PR NEWSWIRE. Xilinx delivers Zynq UltraScale+RFSoC family integrating the RF signal chain for 5 G wireless, cable remote-PHY, and Radar[OL]. [2017-10-03]. https://www.prnewswire.com/news-releases/xilinx-delivers-zynq-ultrascale-rfsoc-family-integrating-the-rf-signal-chain-for-5g-wireless-cable-remote-phy-and-radar-300529479.html
[43]	PAVAN K M P, SUBODH K P. Design and verification of DDR SDRAM memory controller using system verilog for higher coverage[C]// International Conference on Intelligent Computing and Control Systems (ICCS), Madurai, India: IEEE, 2019: 689-694.
[44]	SHENG Z T, WEAVERS P T, TRZASKO J D, et al. Gradient pre-emphasis to counteract first-order concomitant fields on asymmetric MRI gradient systems[J]. Magn Reson Med, 2017, 77(6): 2250-2262. doi: 10.1002/mrm.26315 pmid: 27373901
[45]	BENJAMIN N, PETER B. Simple eddy current compensation by additional gradient pulses[J]. Concepts in Magn Reson Part A, 2018, 47A(1): e21469. doi: 10.1002/cmr.a.2018.47A.issue-1
[46]	NATALIA U, SEBASTIAN L. Advancements in gradient system performance for clinical and research MRI[J]. J Magn Reson Imaging, 2022, 57(1): 57-70. doi: 10.1002/jmri.v57.1
[47]	NIXON T W, MCINTYRE S, ROTHMAN D L, et al. Compensation of gradient-induced magnetic field perturbations[J]. J Magn Reson, 2008, 192(2): 209-217. doi: 10.1016/j.jmr.2008.02.016 pmid: 18329304
[48]	EDELSTEIN W A, KIDANE T K, TARACILA V, et al. Active-passive gradient shielding for MRI acoustic noise reduction[J]. Magn Reson Med, 2005, 53(5): 1013-1017. pmid: 15844137
[49]	SHEN J, NING R P, LIU Y, et al. A method for reducing eddy currents generated by gradient coils[J]. Acta Phys Sin, 2006, 55(6): 3060-3066. doi: 10.7498/aps
	沈杰, 宁瑞鹏, 刘颖, 等. 一种减小梯度线圈产生的涡流的方法[J]. 物理学报, 2006, 55(6): 3060-3066. doi: 10.7498/aps
[50]	刘振友. 1.5 T新生儿自屏蔽磁共振超导磁体研究[D]. 山东: 山东大学, 2021.
[51]	MANSFIELD P, CHAPMAN B. Multishield active magnetic screening of coil structures in NMR[J]. J Magn Reson, 1987, 72(2): 211-233.
[52]	WANG Y, WANG Q, LEI G, et al. An actively shielded gradient coil design for use in planar MRI systems with limited space[J]. Rev Sci Instrum, 2018, 89(9): 095110. doi: 10.1063/1.5043331
[53]	HUANG Q M, ZHENG G. Design method of magnetic resonance biplane self shielding gradient coil based on gradient inductance constraint and graphics processing unit acceleration algorithm[J]. Chin J Med Imaging Technol, 2020, 36(3):453-459.
	黄清明, 郑刚. 基于梯度电感约束和图形处理单元加速算法的磁共振双平面自屏蔽梯度线圈设计方法[J]. 中国医学影像技术, 2020, 36(3): 453-459.
[54]	HE H Y, WEI S F, WANG H X, et al. Matrix gradient coil: current research status and perspectives[J]. Chinese J Magn Reson, 2021, 38(1): 140-153.
	贺红艳, 魏树峰, 王慧贤, 等. 矩阵梯度线圈研究现状与发展趋势[J]. 波谱学杂志, 2021, 38(1): 140-153.
[55]	谢兆媛, 李鲠颖. 具有模拟预加重的梯度波形发生器.第十五届全国波谱学学术会议论文摘要集[C]. 宁波: 2008.
[56]	LIN X C. Research and development of spectrometer, the core of magnetic resonance imaging system[J]. High-Technol& Industrialization, 2013, (12): 70-77.
	林先钗. 磁共振成像系统核心——谱仪的研制开发[J]. 高科技与产业化, 2013, (12): 70-77.
[57]	CHAO D. Development of DC parameter calibration module for digital integrated circuit test system[J]. J Phys Conf Ser, 2023, 2428: 012006. doi: 10.1088/1742-6596/2428/1/012006
[58]	LIU Y, NING R P, LI G Y. A high-precision gradient waveform generator with self triggering function[J]. Chinese J Magn Reson, 2007, 24(4): 433-438.
	刘燕, 宁瑞鹏, 李鲠颖. 一种具有自触发功能的高精度梯度波形发生器[J]. 波谱学杂志, 2007, 24(4): 433-438.
[59]	ZANG F C, LI J Q, WANG H, et al. Research on automatic gradient pre-emphasis adjustment method in magnetic resonance imaging system[J]. Chinese J Magn Reson, 2008, 25(1): 26-32.
	臧凤超, 李建奇, 王鹤, 等. 磁共振成像系统中的自动梯度预加重调节方法研究[J]. 波谱学杂志, 2008, 25(1): 26-32.
[60]	YANG F, XU J C, YAO S Q, et al. Digital eddy current pre-compensation method based on high-precision magnetic resonance gradient card[J]. Chin J Magn Reson Imaging, 2020, 11(5): 354-359.
	杨帆, 徐俊成, 姚守权, 等. 基于高精度磁共振梯度卡的数字涡流预补偿方法[J]. 磁共振成像, 2020, 11(5): 354-359.
[61]	华东师范大学. 一种梯度预补偿波形数字分辨率提高方法及装置: CN201911221339.3[P]. 2021-10-12.
[62]	WANG H, NING R P, LIU Y, et al. A method for eliminating gradient jitter in the design of a digital MRI spectrometer system[J]. Chinese J Magn Reson, 2008, 25(1): 39-45.
	王鹤, 宁瑞鹏, 刘燕, 等. 在数字化MRI谱仪系统设计中消除梯度抖动的方法[J]. 波谱学杂志, 2008, 25(1): 39-45.
[63]	李钰. 具有预失真功能的梯度波形发生器[D]. 华东师范大学, 2015.
[64]	LI Y, HU Y, LUO Q. Gradient waveform generator with pre-distortion function[J]. Acta of electronic meas and instrum, 2015, 29(9): 1374-1381.
	李钰, 胡坤, 罗庆, 等具有预失真功能的梯度波形发生器[J]. 电子测量与仪器学报, 2015, 29(9): 1374-1381.
[65]	SONG R, HE Y F, ZHANG B. A measurement method for eddy current field of magnet in permanent magnetic resonance[J]. Chinese J Magn Reson, 2018, 35(1): 52-59.
	宋瑞, 何砚发, 张波. 一种永磁磁共振中磁体涡流场的测量方法[J]. 波谱学杂志, 2018, 35(1): 52-59.
[66]	WANG Y, HE Y, KANG Y, et al. Automatic gradient pre-emphasis adjustment for permanent MRI system[J]. J Med Imaging Health Inform, 2019, 9(9): 1966-1970. doi: 10.1166/jmihi.2019.2823

运算方式	并行	串行
锁相环	1/4	1/4
逻辑寄存器	11654	10261
嵌入式9位元素乘法器	126	54

厂家（3.0 T）	西门子		飞利浦		通用电气		联影
型号	Skyra	Prisma	Ingenia	Elition	Architect	Premier	uMR 780	uMR 790
最大梯度场强 /（mT/m）	45	80	45	45	44	80	42	100
最大梯度分辨率/（T/m/s）	200	200	200	220	200	200	220	200
厂家（0.5 T）	贝斯达		鑫高益		SternMed		Paramed
型号	BTI-050		OPER-0.5		Marcom		OpenMR
最大梯度场强 /（mT/m）	25		24		25		20
最大梯度分辨率/（T/m/s）	75		70		/		33.3

实现方法	原理	关键技术	分类	优势和劣势
抗涡流板法	由高电阻率材料制成，可以在一定程度上减小涡流	高电阻率材料	/	对厚度的要求严格，设计制作成本高；由于加工工艺限制，板材边缘附近仍会产生涡流^[49]
自屏蔽梯度线圈法	在梯度线圈的外面加一组电流方向与其相反的线圈，使成像区域的梯度磁场满足设计需要	屏蔽线圈^[48]	有源屏蔽梯度线圈	在相邻金属结构上的磁场泄露少，图像质量更好；占据较大的MRI扫描仪空间，且需要对梯度线圈进行特殊而复杂的设计，增加了线圈的成本和功耗
自屏蔽梯度线圈法	在梯度线圈的外面加一组电流方向与其相反的线圈，使成像区域的梯度磁场满足设计需要	屏蔽线圈^[48]	无源屏蔽梯度线圈
梯度波形预加重法	通过施加电流过驱动的方法，产生一个附加补偿磁场，以补偿涡流引起的伴随场	RC电路+滑动变阻器/数控电位器	模拟预加重法	数控电位器取代滑动变阻器的方法保证了数据的非易失性，且受环境等温度影响小；精度低，灵活性差^[55,56]
梯度波形预加重法	通过施加电流过驱动的方法，产生一个附加补偿磁场，以补偿涡流引起的伴随场	DSP/FPGA结合数学物理方法	数字预加重法	能够提高预加重精度并可以灵活扩展预加重参数的个数；占用了DAC位数^[55,56]