Published online before print February 12, 2004, doi: 10.1161/01.STR.0000116102.15636.65
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Stroke. 2004;35:687.)
© 2004 American Heart Association, Inc.
Original Contributions |
Comparison of Transcranial Brain Tissue Perfusion Images Between Ultraharmonic, Second Harmonic, and Power Harmonic Imaging
From the Department of Clinical Neurosciences, Kyoto Takeda Hospital, Kyoto (T.S.); Department of Neurology, Kyoto Prefectural University of Medicine, Kyoto (N.T., T.M., M.N.); and Medical Engineering Laboratory, Tokyo Jikei University School of Medicine, Tokyo (H.F.), Japan.
Correspondence to Toshiyuki Shiogai, MD, Department of Clinical Neurosciences, Kyoto Takeda Hospital, Minamikinuta-cho 11, Nishinanajo, Shimogyo-ku, Kyoto 600-8884, Japan. E-mail shiogait{at}pop11.odn.ne.jp
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Background and Purpose— To clarify optimal brain tissue perfusion images visualized by transcranial ultrasound harmonic imaging, we compared gray-scale integrated backscatter (IBS) images of new ultraharmonic imaging (UHI) and conventional second harmonic imaging (SHI) with power harmonic imaging (PHI) (harmonic B-mode with harmonic power Doppler images) in 10 patients with and 4 without a temporal skull.
Methods— Using a SONOS 5500 (Philips), we evaluated transient response images taken after a bolus Levovist injection at a horizontal diencephalic plane via temporal windows. Based on transmitting/receiving frequencies (MHz), 4 imaging procedures using an S3 transducer (SHI2.6 [1.3/2.6], UHI [1.3/3.6], PHI2.6 [1.3/2.6], and PHI3.2 [1.6/3.2]) and 2 imaging procedures using an S4 transducer (SHI3.6 [1.8/3.6] and PHI3.6 [1.8/3.6]) were compared in terms of size and location, peak intensity (PI), contrast area demarcation, and background image quality.
Results— In intact skull cases, gray-scale imaging tended to show larger contrast areas than PHI. A large contrast area was most frequently observed in SHI2.6 images, despite there being more high-PI cases in UHI. No contrast area with unclear background was observed in a few cases. In craniectomized cases, all contrast images tended to have large and high PI compared with the intact skull cases. PHI, particularly PHI3.6, demonstrated sharper demarcation and a clearer background than gray-scale imaging.
Conclusions— Transcranial gray-scale SHI using a low receiving frequency of 2.6 MHz is the superior method. PHI identifies contrast area localization better than gray-scale imaging and is particularly suitable for intraoperative and postoperative cases.
Key Words: contrast media • perfusion, brain • imaging techniques • ultrasonography, Doppler, color, transcranial
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Transcranial ultrasonic imaging using contrast-specific phenomena has been tried clinically in evaluation of brain tissue perfusion.1–3 The usefulness for patients with ischemic stroke4–8 and with cognitive impairments9 has been pointed out. The physical properties of microbubbles as echo-contrast agents in an acoustic field depend mainly on acoustic power. At low acoustic emission power, the bubbles act as linear backscatters,10 and to receive these signals, conventional fundamental imaging such as B-mode images on color velocity (CV) or power Doppler (PD) images have been used. Initial experiments have shown that brain tissue perfusion by the fundamental imaging of B-mode11 and PD12 with high-frequency transducers was necessary to insonate via a skull defect. However, fundamental imaging using linear backscatter does not sufficiently visualize brain tissue perfusion. With an increase in emission power, microbubbles start to show resonance phenomena and nonlinear stationary scattering where emitted signals have not only the fundamental frequency (same as the transmitting frequency) but also harmonic components (multiples of the transmitting frequency).13 When the insonated power level is increased, microbubble destruction results in the production of intense transient nonlinear backscatter, referred to as stimulated acoustic emission (SAE).10,14
Using the harmonic or SAE mechanism, visualization of brain tissue perfusion with low-frequency transducers via temporal acoustic windows has been clinically evaluated. Instead of conventional B-mode imaging, gray-scale second harmonic imaging (SHI) using integrated backscatter (IBS),1,3,9 directly integrated radiofrequency signals from brain tissue, or pulse-inversion methods2,5 have been used for transcranial harmonic imaging. Fundamental imaging with SAE such as background B-mode with color-coded imaging of CV7 or PD14 and parameter imaging calculated from time-intensity curve, called contrast burst imaging or time variance imaging,2 have been introduced.
In echocardiography, SHI combined with PD is the most sensitive technique in terms of contrast-to-tissue ratio (CTR)15 and microbubble detection at the perfusion level.13 However, for transcranial application, SHI with PD or power harmonic imaging (PHI) has only been studied experimentally.16,17 Clinical studies outside preliminary observations18 have not been published.
Recently, to increase CTR, new ultraharmonic imaging (UHI) using beyond the second but below the third level of harmonics has been used for contrast echocardiography. UHI produces low precontrast tissue signals, exhibits efficient microbubble destruction with multiple-frame triggering,19 and optimizes postcontrast myocardial opacification. However, transcranial effectiveness has also yet to be clarified.
Until now, an optimal imaging procedure that considers transmitting and receiving frequencies and the effect of temporal skull has not been fully elucidated for transcranial evaluation of brain tissue perfusion. Therefore, to clarify optimal transcranial images by contrast-enhanced harmonic imaging, we compared the gray-scale IBS images of UHI, SHI, and PHI with different transmitting and receiving frequencies in patients with and without a temporal skull.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
We used a SONOS5500 (Philips Medical Systems) ultrasound system with new S3 and old S4 ultraband transducers that cover frequency ranges from 1.3 to 3.6 and 1.8 to 3.6 MHz, respectively. The transducer was fixed securely on the temporal acoustic window during contrast studies. The investigation depth was 12 cm, and the focus was 6 cm (12 cm in 1 PHI case). The harmonic imaging was performed on the transtemporal axial diencephalic plane (landmarks were the third and anterior horns of lateral ventricles) involving the temporal lobe (TL), basal ganglia (BG), and thalamus (Th) that were confirmed by fundamental B-mode and CV images before and after the contrast studies. Sixty-two transient response images triggered by ECG were acquired at 1 image per 2 cardiac cycles (sampling intervals, 1.7±0.3 seconds) after a bolus 7-mL (300-mg/mL) injection of Levovist (Schering) through an antecubital vein catheter, and the data storing on a magneto-optical disk were verified.
In gray-scale IBS images of UHI and SHI, the mechanical index, system gain, and compression were 1.5 (or 1.6), 75, and 70 (or 75), respectively. In PHI, a harmonic PD image is superimposed on the background harmonic B-mode image with parameters identical to the IBS images. The PD image settings were as follows: color gain, 40%; threshold, 10; pulse repetition frequency, 3.9 kHz; and filter, 4. Quantification was performed by T-INT mode and analyzed by acoustic densitometry20 for IBS images and Quanti-Con (EchoTech Co) for PHI images.
On the basis of transmitting/receiving frequencies (MHz), 6 imaging procedures were compared; the principal study involved 4 imaging procedures using an S3 transducer (IBS images of SHI2.6 [1.3/2.6] and UHI [1.3/3.6], PHI2.6 and PHI3.2 [both B-mode images of 1.6/3.2, PD images 1.3/2.6 and 1.6/3.2]), and the previous pilot study concerned 2 imaging procedures using an S4 transducer (SHI3.6 [(1.8/3.6] and PHI3.6 [1.8/3.6]). The 6 imaging techniques were evaluated in terms of maximum size (diameter) and location of contrast area, maximum peak intensity (PI) based on time-intensity curve analysis in the three regions of interest (TL, BG, and Th), sharpness of contrast area demarcation, and background image quality (clarity).
The subjects were fourteen stable patients in rehabilitation wards with open temporal acoustic windows previously confirmed by fundamental CV imaging. The ten patients with an intact temporal skull were compared with four patients who had received temporal craniectomy. The previous pilot study of 2 imaging procedures (SHI3.6 and PHI3.6) was performed on the same day. The following principal study of 4 imaging procedures (SHI2.6, UHI, PHI2.6, and PHI3.2) was performed on the same day 0.5 to 11 months (mean, 3 months) after the pilot study. No differences in patients’ neurological conditions and pathologies on imaging studies were confirmed before the principal studies. Patients’ demographics are presented in Table 1.
|
Informed consent was obtained from patients and/or family members before the study. We used Student’s t and 
2 tests; statistical significance was set at P<0.05.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Intact Temporal Skull Cases
Size and Location of Contrast Area
A large contrast area (5 to <7 cm in diameter) and wider range of locations involving the TL, BG, and Th were most frequently observed in SHI2.6 images compared with UHI and SHI3.6 images (Table 2). In PHI images, there was a tendency toward smaller contrast areas in PHI3.2 than in PHI2.6 or PHI3.6. Gray-scale imaging tended to have larger (P<0.01) and wider range of location (P<0.001) of contrast areas than PHI.
|
Contrast Intensity
There were more cases of high PI (>10 dB) in UHI than other images and a tendency toward low PI in PHI3.2 and PHI3.6 (P<0.05).
Contrast Area Demarcation and Background Image Clarity
The frequency of sharp demarcation was higher in SHI2.6, UHI, PHI2.6, and PHI3.6 cases than in SHI3.6 and PHI3.2 cases. There was greater difficulty in identifying the anatomical localization of the contrast area caused by unclear background images, particularly in UHI, SHI3.6, and PHI3.2 (Figure 1b, 1c, and 1e). No contrast area associated with extremely unclear background images was observed in SHI2.6 and UHI in 2 cases, SHI3.6 in 3 cases, and PHI3.2 and PHI3.6 in 1 case each, despite small contrast spots in the near field from the transducer being identified in PHI2.6 cases.
|
Craniectomized Cases
The contrast area of all images in all craniectomized cases but 1 became larger than those in intact skull cases (Table 3). The contrast effects in the near field were more obvious in PHI than in gray-scale imaging (Figure 2). Size and location of contrast areas tended to increase in images using higher receiving frequencies, particularly in PHI3.6 (Figure 2f), but resulted in no significant differences. In all images, high PI (>10 dB) was observed in all but 1 case. Contrast area demarcation was always sharper in PHI and UHI compared with SHI (P<0.01). However, background image was unclear in all UHI and 2 SHI2.6 cases (P=0.06).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Gray-Scale Harmonic Imaging
A clearer distinction between tissue and contrast bubbles, which is expressed by CTR,15,21 increases as a function of the harmonic frequency order (second, third, and higher harmonics). Imaging based on higher harmonics shows a higher CTR, which is significantly influenced by precontrast tissue signals, than imaging at a lower or fundamental frequency. UHI rejects tissue signals by receiving signals beyond the second but below the third harmonics. In myocardial contrast echocardiography, UHI produces low precontrast tissue signals and efficient postcontrast myocardial opacification.19
In our patients with a temporal skull, transcranial UHI was not always superior to SHI. Although a high PI in UHI was observed more frequently than in SHI, a large contrast area was more frequently identified in SHI2.6 than in UHI. The ultraharmonics scattered from bubbles are usually low-energy components compared with second harmonics.21 Attenuation of ultrasonic signals by temporal skull depends on both the transmitting and receiving frequencies. Higher-frequency signals are more attenuate than lower-frequency signals. Therefore, ultraharmonic signals of a higher receiving frequency of 3.6 MHz are more attenuated than second harmonic signals of 2.6 MHz. Furthermore, the tissue signal rejection itself in increased CTR resulted in a further drawback in UHI, whereby the background image was always unclear in patients both with and without a temporal skull.
In transcranial harmonic gray-scale imaging, aside from the pulse-inversion method,2,5 harmonic IBS1,3,4,9,18 has been used more frequently than harmonic B-mode.6 In terms of background image, harmonic B-mode in PHI tended to identify anatomical localization more easily than harmonic IBS in gray-scale imaging, particularly in skull defect cases (Table 3 and Figure 2). However, myocardial contrast enhancement using open-chest dogs increased equally with either B-mode or IBS imaging.22 Further clarifications are necessary to determine superiority.
Gray-Scale Harmonic Imaging and PHI
In echocardiography, SHI with PD (or PHI) is the most sensitive technique.13,15 In our intact skull cases, however, the contrast areas of PHI images using Levovist were of a smaller size and lower PI than gray-scale imaging of UHI and SHI. The same superiority of transcranial gray-scale SHI was reported in serial experiments using dogs with 2 different echo-contrast agents, Optizon16 and Sonazoid.17 This inferiority of PHI is probably due to acoustic pressure attenuation by the skull. PHI requires destruction of bubbles in the microcirculation to generate signals. Bubble destruction requires application of a higher ultrasound power compared with the harmonic response. However, harmonic gray-scale imaging will display bubbles if they have significant harmonic response without destruction.16
The distribution of acoustic pressures in sector scans is inhomogeneous, so microbubbles are barely detectable in the near field. In transcranial harmonic PD, there were no signal increases in the posterior part of the brain.16,17 Despite the use of parameter imaging derived from transcranial gray-scale SHI,3,8 it was difficult to delineate contrast areas in the anterior and posterior borders of the ultrasound section. However, in our craniectomized cases, the contrast effects in the near field were more obvious in PHI than in gray-scale IBS imaging (Figure 2). Furthermore, there was sharp contrast area demarcation in all cases and clear background in all cases but 1. These results indicate that PHI would be suitable for skull defect cases. However, the relationship between high acoustic pressure with microbubble contrast agents and local microvascular damage should be taken into consideration.23
Contrast-specific SAE signals are detectable by transcranial CV or PD imaging without the use of harmonic technology.14 Contrast burst imaging and time variance imaging using the SAE mechanism improved diagnostic sensitivity compared with gray-scale pulse-inversion SHI with transmitting and receiving frequencies of 1.3 and 2.6 MHz.2 The contrast signals are display color coded, but the tissue signal is suppressed. Therefore, it is necessary to change from contrast burst imaging image to harmonic B-mode image to decide anatomical localization before contrast infusion and repositioning of the transducer as a result of patient movement.2 In addition, parameter images relating to PI were introduced, although the information from the background anatomical localization proved unsatisfactory. To overcome this problem, color-coded parameter imaging derived from gray-scale SHI overlaid on MRI has been introduced.3 However, this offline analysis would not be suitable for acute stroke patients. In contrast, conventional or pulse-inversion SHI and background harmonic B-mode of PHI display tissue harmonic images and facilitate permanent anatomic orientation. In PHI additionally, contrast PD images on the background harmonic B-mode image can immediately be switched on and off.
Despite the use of low receiving frequencies, images with no transcranial contrast were almost always associated with an unclear background and were more frequently identified in gray-scale imaging than PHI. The probable reason is that color-coded imaging such as PHI identifies even a small contrast area more easily than gray-scale imaging.
For Optimal Perfusion Imaging
To overcome the major drawback of transcranial ultrasound attenuation, further developments in ultrasound equipment and contrast agents are essential. However, the importance of appropriate frequency selection and beam configuration for transcranial imaging has been pointed out.24 Furthermore, microbubble destruction is more evident at lower frequencies and higher acoustic powers.25 Therefore, subharmonic imaging using a frequency range of around 0.5 to 1 MHz24 would probably be suitable for transcranial perfusion imaging because of the much smaller attenuation of scattered subharmonic signals26 and the high CTR.27
Finally, for intact temporal skull cases, gray-scale SHI using a low receiving frequency of 2.6 MHz is the superior method. Therefore, clinical application for acute stroke cases, particularly interventional occlusion and recanalization of major arteries, is recommended. Compared with gray-scale imaging, PHI can more easily identify contrast area localization, especially in skull defect cases, and therefore is more suitable for intraoperative and postoperative follow-up.
Received June 26, 2003; revision received August 30, 2003; accepted November 4, 2003.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
- Postert T, Muhs A, Meves S, Federlein J, Przuntek H, Buttner T. Transient response harmonic imaging: an ultrasound technique related to brain perfusion. Stroke. 1998; 29: 1901–1907.
[Abstract/Free Full Text] - Postert T, Hoppe P, Federlein J, Helbeck S, Ermert H, Przuntek H, Buttner T, Wilkening W. Contrast agent specific imaging modes for the ultrasonic assessment of parenchymal cerebral echo contrast enhancement. J Cereb Blood Flow Metab. 2000; 20: 1709–1716.[CrossRef][Medline] [Order article via Infotrieve]
- Wiesmann M, Seidel G. Ultrasound perfusion imaging of the human brain. Stroke. 2000; 31: 2421–2425.
[Abstract/Free Full Text] - Postert T, Federlein J, Weber S, Przuntek H, Buttner T. Second harmonic imaging in acute middle cerebral artery infarction: preliminary results. Stroke. 1999; 30: 1702–1706.
[Abstract/Free Full Text] - Meairs S, Daffertshofer M, Neff W, Eschenfelder C, Hennerici M. Pulse-inversion contrast harmonic imaging: ultrasonographic assessment of cerebral perfusion. Lancet. 2000; 355: 550–551.[CrossRef][Medline] [Order article via Infotrieve]
- Federlein J, Postert T, Meves S, Weber S, Przuntek H, Buttner T. Ultrasonic evaluation of pathological brain perfusion in acute stroke using second harmonic imaging. J Neurol Neurosurg Psychiatry. 2000; 69: 616–622.
[Abstract/Free Full Text] - Schlachetzki F, Hoelscher T, Dorenbeck U, Greiffenberg B, Marienhagen J, Ullrich OW, Bogdahn U. Sonographic parenchymal and brain perfusion imaging: preliminary results in four patients following decompressive surgery for malignant middle cerebral artery infarct. Ultrasound Med Biol. 2001; 27: 21–31.[CrossRef][Medline] [Order article via Infotrieve]
- Meyer K, Wiesmann M, Albers T, Seidel G. Harmonic imaging in acute stroke: detection of a cerebral perfusion deficit with ultrasound and perfusion MRI. J Neuroimaging. 2003; 13: 166–168.[CrossRef][Medline] [Order article via Infotrieve]
- Shiogai T, Uebo C, Makino M, Mizuno T, Nakajima K, Furuhata H. Acetazolamide vasoreactivity in vascular dementia and persistent vegetative state evaluated by transcranial harmonic perfusion imaging and Doppler sonography. Ann N Y Acad Sci. 2002; 977: 445–453.
[Abstract/Free Full Text] - De Jong N. Physics of microbubble scattering. In: Nanda NC, Schlief R, Goldberg BB, eds. Advances in Echo Imaging Using Contrast Enhancement. 2nd ed. Dubai, Saudi Arabia: Kluwer Academic Publishers; 1997: 39–64.
- Itoh T, Matsumoto M, Uchimoto R, Takahashi N, Niwa K, Handa N, Tsukamoto Y, Miyazawa T, Kamada T. Perfusion imaging of the brain by B-mode ultrasonography: an experimental study in rabbits. Stroke. 1995; 26: 2353–2356.
[Abstract/Free Full Text] - Greenberg RS, Taylor GA, Stapleton JC, Hillsley CA, Spinak D. Analysis of regional cerebral blood flow in dogs, with an experimental microbubble-based US contrast agent. Radiology. 1996; 201: 119–123.
[Abstract/Free Full Text] - Burns PN. Instrumentation for contrast echocardiography. Echocardiography. 2002; 19: 241–258.[CrossRef][Medline] [Order article via Infotrieve]
- Pohl C, Tiemann K, Schlosser T, Becher H. Stimulated acoustic emission detected by transcranial color Doppler ultrasound: a contrast-specific phenomenon useful for the detection of cerebral tissue perfusion. Stroke. 2000; 31: 1661–1666.
[Abstract/Free Full Text] - Frinking PJ, Bouakaz A, Kirkhorn J, Ten Cate FJ, de Jong N. Ultrasound contrast imaging: current and new potential methods. Ultrasound Med Biol. 2000; 26: 965–975.[CrossRef][Medline] [Order article via Infotrieve]
- Seidel G, Algermissen C, Christoph A, Katzer T, Kaps M. Visualization of brain perfusion with harmonic gray scale and power Doppler technology: an animal pilot study. Stroke. 2000; 31: 1728–1734.
[Abstract/Free Full Text] - Meyer K, Seidel G, Algermissen C. Harmonic imaging of the brain parenchyma in a dog model following NC100100 (Sonazoid) bolus injection. J Neuroimaging. 2002; 12: 35–41.[Medline] [Order article via Infotrieve]
- Shiogai T, Tsunezuka C, Koshimura M. Takayasu N, Nagakane Y, Nakase T, Makino M, Nakajima K, Furuhata H. Comparative visualization of brain tissue perfusion in various neurological disorders utilizing contrast-enhanced harmonic imaging in integrated backscatter and power Doppler methods. Cerebrovasc Dis. 2000; 10 (suppl 1): 44.Abstract.[CrossRef][Medline] [Order article via Infotrieve]
- Kuersten B, Murthy TH, Li P, Liu Z, Locricchio E, Baisch C, Rafter P, Vannan M. Ultraharmonic myocardial contrast imaging: in vivo experimental and clinical data from a novel technique. J Am Soc Echocardiogr. 2001; 14: 910–916.[CrossRef][Medline] [Order article via Infotrieve]
- D’Sa A. Acoustic Densitometry. Coopertino, Calif: Hewlett-Packard Co; white paper 1999.
- Bouakaz A, Frigstad S, Ten Cate FJ, de Jong N. Improved contrast to tissue ratio at higher harmonics. Ultrasonics. 2002; 40: 575–578.[CrossRef][Medline] [Order article via Infotrieve]
- Rovai D, Janerot-Sjoberg B, Nagy A, Marini C, Burchielli S, Castellari M, Morales MA, Giovanna Trivella M, Ostensen J, Distante A, L’Abbate A. Myocardial perfusion abnormalities by intravenous administration of the contrast agent NC100100 in an experimental model of coronary artery thrombosis and reperfusion. Echocardiography. 1998; 15: 731–740.[Medline] [Order article via Infotrieve]
- Ay T, Havaux X, Van Camp G, Campanelli B, Gisellu G, Pasquet A, Denef JF, Melin JA, Vanoverschelde JL. Destruction of contrast microbubbles by ultrasound: effects on myocardial function, coronary perfusion pressure, and microvascular integrity. Circulation. 2001; 104: 461–416.
[Abstract/Free Full Text] - Fry FJ, Barger JE. Acoustical properties of the human skull. J Acoust Soc Am. 1978; 63: 1576–1590.[CrossRef][Medline] [Order article via Infotrieve]
- Wei K, Skyba DM, Firschke C, Jayaweera AR, Lindner JR, Kaul S. Interactions between microbubbles and ultrasound: in vitro and in vivo observations. J Am Coll Cardiol. 1997; 29: 1081–1088.[Abstract]
- Forsberg F, Shi WT, Goldberg BB. Subharmonic imaging of contrast agents. Ultrasonics. 2000; 38: 93–98.[CrossRef][Medline] [Order article via Infotrieve]
- Krishna PD, Shankar PM, Newhouse VL. Subharmonic generation from ultrasonic contrast agents. Phys Med Biol. 1999; 44: 681–694.[CrossRef][Medline] [Order article via Infotrieve]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||