Analysis of spatial dependence of acoustic noise transfer function in magnetic resonance imaging

Transcription

1 Analysis of spatial dependence of acoustic noise transfer function in magnetic resonance imaging Award: Magna Cum Laude Poster No.: C-1988 Congress: ECR 2014 Type: Scientific Exhibit Authors: T. Hamaguchi, T. Miyati, T. Matsushita, N. Ohno ; Kanazawa/ JP, Kyoto/JP Keywords: Safety, MR, MR physics, Quality assurance DOI: /ecr2014/C-1988 Any information contained in this pdf file is automatically generated from digital material submitted to EPOS by third parties in the form of scientific presentations. References to any names, marks, products, or services of third parties or hypertext links to thirdparty sites or information are provided solely as a convenience to you and do not in any way constitute or imply ECR's endorsement, sponsorship or recommendation of the third party, information, product or service. ECR is not responsible for the content of these pages and does not make any representations regarding the content or accuracy of material in this file. As per copyright regulations, any unauthorised use of the material or parts thereof as well as commercial reproduction or multiple distribution by any traditional or electronically based reproduction/publication method ist strictly prohibited. You agree to defend, indemnify, and hold ECR harmless from and against any and all claims, damages, costs, and expenses, including attorneys' fees, arising from or related to your use of these pages. Please note: Links to movies, ppt slideshows and any other multimedia files are not available in the pdf version of presentations. Page 1 of 25

2 Aims and objectives Acoustic noise during magnetic resonance imaging (MRI) scan can not only cause anxiety, stress, and annoyance but also cause a temporary or permanent shift in the hearing threshold [1]. Studies have described that this sound affects the results of functional MRI [2]. In addition, the noise interferes with verbal communication and pediatric imaging. The noise is primarily caused by Lorenz forces (F) induced in the scanner while electric currents (I) are switched in the gradient coils within the static magnetic field (B0) (figure 1). Fig. 1: Acoustic noise during MRI scan is primarily caused by Lorenz forces (F) induced in the scanner while electric currents (I) are switched in the gradient coils within the static magnetic field (B0). Acoustic noise transfer function can evaluate the noise properties of an MRI system itself, independently of pulse sequences and imaging parameters [3]. The acoustic noise transfer function may vary depending on the measurement positions (corresponding to the patient positions) as well as the MRI systems. Therefore, the present study aimed to evaluate the spatial dependence of gradient-pulse-to-acoustic-noise transfer function (GPAN-TF) among different MRI systems. Images for this section: Page 2 of 25

3 Fig. 1: Acoustic noise during MRI scan is primarily caused by Lorenz forces (F) induced in the scanner while electric currents (I) are switched in the gradient coils within the static magnetic field (B0). Page 3 of 25

4 Methods and materials We measured sound pressure levels induced by gradient impulses in a 3.0T superconducting MRI system (Signa Excite HDxt, General Electric Healthcare, Wisconsin, USA) and a 0.4-T permanent magnet open MRI system (APERTO Eterna, Hitachi medical corporation, Tokyo, Japan) (figure 2). Measurements were performed using a precision integrating sound level meter (ISLM)(NL-18, RION Co., Ltd., Tokyo, Japan) connected to an electret condenser microphone. Fig. 2: a) 3.0-T superconducting MRI system (Signa Excite HDxt, General Electric Healthcare, Wisconsin, USA), b) 0.4-T permanent magnet open MRI system (APERTO Eterna, Hitachi medical corporation, Tokyo, Japan). The measurement positions inside the bore in both the systems were 143 and 75 points, respectively (figure 3). A simple trapezoidal pulse was applied to X, Y, and Z gradient coils, as shown in figure 4. The gradient pulse amplitude, rise and fall times, and gradient pulse on time are summarized in Table 1. Within the manufacturer's pulse sequence design limitations, the rise and fall times and gradient pulse on time were the minimum allowed, whereas the amplitude was the maximum allowed. Acoustic noise waveforms induced by a single gradient pulse were digitalized and sampled at 96 khz, following which data were acquired in the frequency domain using a computational tool, including Fourier transforms. To measure the peak sound pressure levels, a microphone was Page 4 of 25

5 directed parallel to the patient axis and in a horizontal plane, without a patient or phantom on the patient table (figure 5). Page 5 of 25

6 Fig. 3: The measurement positions inside the bore in the a) superconducting MRI system (143 points) and b) permanent magnet open MRI system (75 points). Fig. 4: Gradient pulse were excited for each gradient. Pulse sequence was programmed to excite X, Y, and Z coil, respectively. The gradient pulse amplitude, rise and fall times, and gradient pulse on time are summarized in Table 1. Page 6 of 25

7 Table 1: Characteristic of narrow trapezoidal input waveform Fig. 5: Experimental set-up for acoustic noise measurement. The MRI scanner was confirmed to be a linear system between acoustic noise and gradient amplitude [4]. We also conducted preliminary tests to check the linearity in both the scanners (figure 6). As shown in figure 7, we used the deconvolution process to calculate GPAN-TFs [µpa/(mt/m)] for the various spatial measurement positions inside each gradient coil (i.e., X-, Y-, and Z-axes). Page 7 of 25

8 Fig. 6: Relation between gradient amplitude and acoustic noise. A strong positive correlation was noted. There is a linear dependency between acoustic noise and gradient amplitude. Page 8 of 25

9 Fig. 7: Calculation of GPAN-TF in each gradient coil (X, Y, and Z-axis), by the deconvolution process. Images for this section: Fig. 2: a) 3.0-T superconducting MRI system (Signa Excite HDxt, General Electric Healthcare, Wisconsin, USA), b) 0.4-T permanent magnet open MRI system (APERTO Eterna, Hitachi medical corporation, Tokyo, Japan). Page 9 of 25

10 Fig. 3: The measurement positions inside the bore in the a) superconducting MRI system (143 points) and b) permanent magnet open MRI system (75 points). Page 10 of 25

11 Fig. 4: Gradient pulse were excited for each gradient. Pulse sequence was programmed to excite X, Y, and Z coil, respectively. The gradient pulse amplitude, rise and fall times, and gradient pulse on time are summarized in Table 1. Page 11 of 25

12 Table 1: Characteristic of narrow trapezoidal input waveform Fig. 5: Experimental set-up for acoustic noise measurement. Page 12 of 25

13 Fig. 6: Relation between gradient amplitude and acoustic noise. A strong positive correlation was noted. There is a linear dependency between acoustic noise and gradient amplitude. Page 13 of 25

14 Fig. 7: Calculation of GPAN-TF in each gradient coil (X, Y, and Z-axis), by the deconvolution process. Page 14 of 25

15 Results The acoustic noise magnitude induced by the gradient impulse varied from 93.5 db to db (3.0 T) and from 76.9 db to db (0.4 T), according to the measurement positions. Figure 8 shows the spatial distribution of GPAN-TFs of X, Y, and Z gradient coils in a superconducting MRI (3.0 T). GPAN-TF of the X coil reached its peak at ±X sides from the isocenter (figure 9). GPAN-TF of the Y coil had a maximum value at a higher position than the isocenter (figure 10a). GPAN-TF of the Z coil showed bimodal peaks on the patient axis (figure 10b). Fig. 8: Spatial distribution of GPAN-TFs in a superconducting MRI (3.0 T). Page 15 of 25

16 Fig. 9: GPAN-TF of the X coil reached its peak at ±X sides from the isocenter. Fig. 10: a) GPAN-TF of the Y coil had a maximum value at a higher position than the isocenter (Z=0). b) GPAN-TF of the Z coil showed bimodal peaks on the patient axis (Y=0). Figure 11 shows the spatial distribution of GPAN-TFs of X, Y, and Z gradient coils in a permanent magnet open MRI (0.4 T). Maximum values of GPAN-TFs for the X and Y coils were present in ±X and ±Y directions from the isocenter, respectively (figure 12). Page 16 of 25

17 The noise generated by the Z coil was exceedingly smaller than that generated by the X and Y coils at all the measurement positions. Fig. 11: Spatial distribution of GPAN-TFs in a permanent magnet open MRI (0.4 T). Fig. 12: Maximum values of GPAN-TFs for the a) X and b) Y coils were present in ±X and ±Y directions from the isocenter, respectively. Figure 13 shows the shapes and designs of the gradient coils in the two MR systems. The spatial distribution of GPAN-TFs by multipoint measurement revealed the properties on the basis of the structural differences in each MRI scanner type and their respective Page 17 of 25

18 gradient coils. This result shows that the "hotspot" area of GPAN-TFs may be associated with acoustic noise sources. It may be able to effectively reduce the noise because the propagating sound from the gradient coils to the "hotspot" was shielded. Fig. 13: Shapes and designs of the gradient coils in the a) 3.0-T superconducting MRI system and b) 0.4-T permanent magnet open MRI system. Page 18 of 25

19 GPAN-TFs at the isocenter were not always higher than those at the other measurement positions (figure 14). A traditional measurement method for the noise during MRI was standardized only under the isocenter and the patient axis [5]. Therefore, the patient would be exposed to a more intense noise than initially planned. Fig. 14: Comparison of boxplot distribution for GPAN-TFs at all measurement positions in the 3.0-T (left) and 0.4-T (right) MRI systems. In particular, at a high frequency (>1000 Hz), GPAN-TF of the superconducting MRI scanner (3.0 T) was significantly higher than that of the permanent magnet open MRI scanner (0.4 T) (P < 0.001, unpaired t-test) (figure 15). The acoustic noise transfer for the high frequency content may be attributed to the increase in the uncomfortable acoustic noise of the high static field scanner. Page 19 of 25

20 Fig. 15: Spectra of GPAN-TFs in the a) 3.0-T and b) 0.4-T MRI scanner at isocenter. Images for this section: Fig. 8: Spatial distribution of GPAN-TFs in a superconducting MRI (3.0 T). Page 20 of 25

21 Fig. 9: GPAN-TF of the X coil reached its peak at ±X sides from the isocenter. Fig. 10: a) GPAN-TF of the Y coil had a maximum value at a higher position than the isocenter (Z=0). b) GPAN-TF of the Z coil showed bimodal peaks on the patient axis (Y=0). Page 21 of 25

22 Fig. 11: Spatial distribution of GPAN-TFs in a permanent magnet open MRI (0.4 T). Fig. 12: Maximum values of GPAN-TFs for the a) X and b) Y coils were present in ±X and ±Y directions from the isocenter, respectively. Page 22 of 25

23 Fig. 13: Shapes and designs of the gradient coils in the a) 3.0-T superconducting MRI system and b) 0.4-T permanent magnet open MRI system. Page 23 of 25

24 Fig. 14: Comparison of boxplot distribution for GPAN-TFs at all measurement positions in the 3.0-T (left) and 0.4-T (right) MRI systems. Fig. 15: Spectra of GPAN-TFs in the a) 3.0-T and b) 0.4-T MRI scanner at isocenter. Page 24 of 25

25 Conclusion Multipoint analysis using GPAN-TFs facilitates detailed evaluation of the spatial dependence of acoustic noise among different MRI systems. Personal information References [1] McJury M, Shellock F. Auditory noise associated with MR procedures: a review. J Magn Reson Imaging 2000; 12: [2] Zhang NY, Zhu XH, Chen W. Influence of gradient acoustic noise on fmri response in the human visual cortex. Magn Reson Med 2005; 54: [3] Hamaguchi T, Miyati T, et al. Acoustic noise transfer function in clinical MRI: a multicenter analysis. Acad Radiol, 2011; 18: [4] Hedeen R, Edelstein W. Characterization and prediction of gradient acoustic noise in MR imagers. Magn Reson Med 1997; 37:7-10. [5] National Electrical Manufacturers Association. Acoustic noise measurement procedure for diagnostic magnetic resonance imaging devices. NEMA Standards Publication MS Page 25 of 25