| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Department of Radiology, University of Kansas Medical Center, 3901 Rainbow
Blvd., Kansas City, KS 66160-7234.
2
Siemens Medical Ultrasound Group, 22010 S.E. 51st St., Issaquah, WA
98029-7002.
Received June 12, 2000;
accepted after revision December 4, 2000.
Address correspondence to S. J. Rosenthal.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Two hundred randomly selected patients were scanned with both fundamental and tissue harmonic methods on a sonography system. The following transducers were used: 2.5 MHz, 20 mm phased array; 3 MHz, 40 mm curved array; 6.5 MHz, 13 mm curved array transvaginal; and 7.5 MHz, 40 mm linear. Operators evaluated visualization of normal and pathologic tissues by tissue harmonic versus fundamental imaging using scores ranging from 1 for much worse visualization to 5 for much better visualization. They also assessed the overall utility of tissue harmonic imaging in the diagnosis of the patient's condition. The studies were saved on magnetooptical disc and were independently reviewed by one of the authors.
RESULTS. Tissue harmonic imaging was helpful for all types of examination. Tissue harmonic imaging improved visualization of normal tissue in 49% of the cases and pathologic tissue in 73% of the cases. Tissue harmonic imaging was found to be diagnostically helpful in 43% of the cases and essential to the diagnosis in 6% of the cases.
CONCLUSION. Tissue harmonic imaging significantly improves visualization of both normal and pathologic tissues and its selective use has major diagnostic utility in a wide variety of clinical applications.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Images formed from the harmonic echoes have a somewhat different appearance from conventional scans produced by echoes of the fundamental transmitted beam. Because the harmonic beam is amplified rather than attenuated by passage through tissue, few harmonic echoes are produced by the skin and subcutaneous tissues. This results in a dramatic decrease in image aberration and deterioration caused by the body wall. The signal-to-noise ratio benefits not only from the dramatically decreased body wall factor, but also from a significant reduction in artifacts caused by reverberation and side lobes. These artifacts are weaker than the fundamental beam, so they produce much weaker harmonic echoes. Penetration can be improved using a low fundamental transmitted pulse and imaging at the higher harmonic frequency. More harmonics are produced at the center of a broadband pulse than in the periphery. This gives a narrower beam, which improves spatial resolution and further reduces side lobes, resulting in improved contrast resolution [1,2,3].
As much of the fundamental sonographic signal as possible must be removed to make these theoretical harmonic improvements a clinical reality. This can be done by either frequency-based or phase inversion methods. Previously published clinical studies have used the frequency-based method [3,4,5,6]. This technique uses a narrow transmit frequency bandwidth. A high-pass or narrow band-pass filter is critical to suppress the fundamental signal and receive the second harmonic at twice the transmitted frequency. Any parts of the harmonic signal that overlap the fundamental are lost. This method has been successful in improving images of deep abdominal and pelvic structures and of obese patients.
The studies we report in this article were performed using an alternative phase inversion technology. With this method, two consecutive wide-bandwidth pulses are transmitted. The phase of the second is inverted 180° relative to the first. The returning echoes are summed. The linear responses from the fundamental beam and odd harmonic echoes are suppressed. Even harmonic echoes, especially the second, are amplified. This amplification allows wideband filtering to more effectively eliminate the fundamental signal with less harmonic loss. Multifrequency capacity is maintained. The technology has been implemented on a variety of clinical transducers (Figs. 1 and 2A,2B).
|
|
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
All examinations were performed on a SONOLINE Elegra scanner (Siemens Ultrasound Group, Issaquah, WA). Harmonic scans used "Ensemble Tissue Harmonic Imaging," (Siemens Ultrasound Group) a proprietary form of phase inversion tissue harmonic imaging. The following multifrequency transducers were used: 2.5 MHz, 20 mm phased; 3.0 MHz, 40 mm curved; 6.5 MHz, 13 mm transvaginal; and 7.5 MHz, 40 mm linear transducers. Each transducer has three to five fundamental and harmonic frequencies. Tissue harmonic imaging was activated by a single button control. The operators then optimized frequency, time-gain compensation, and system gain to maximize resolution of the structures of interest.
All examinations were performed by one of six experienced, registered sonographers. All examinations were stored on magnetooptical discs and independently reviewed by a radiologist with more than 20 years of sonography experience. Whether tissue harmonic imaging or fundamental scanning was performed first was randomized solely at the discretion of the operator. With the equipment used, the fundamental frequency comes up first when any transducer is initialized, but tissue harmonic imaging was used first in 53% of the examinations.
Registered sonographers scored all examinations, evaluating patient body habitus, examination difficulty, visualization of normal tissues, visualization of pathologic tissues, and the overall diagnostic utility of tissue harmonic imaging. The evaluations and images were also independently reviewed by a radiologist. We gave greater weight to showing pathology when evaluating the diagnostic utility of studies with both normal and pathologic tissues. In cases of discrepancy between the radiologist's score and sonographer's score, they reviewed the soft copy together and the final score was obtained after discussion and mutual agreement.
We used a subjective scoring system to evaluate visualization of normal tissue, pathologic tissue, and overall diagnostic utility. We graded according to the following 5-point classification: 1, tissue harmonic imaging was much worse than fundamental; 2, tissue harmonic imaging was worse than fundamental; 3, tissue harmonic imaging was diagnostically equal to fundamental; 4, tissue harmonic imaging was better than fundamental; and 5, tissue harmonic imaging was essential for the diagnosis. Average quality scores were calculated for normal tissue, pathologic tissue, and overall tissue harmonic imaging utility. In each category, an average score was calculated based on the 5-point classification. Standard deviation and confidence level (p = 0.05) were also calculated. In the three main categories, tissue harmonic imaging scores were summed for all examinations and divided into abdominal, obstetric, pelvic, small parts, and other (superficial vascular and musculoskeletal) groups.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
|
Tissue harmonic imaging was more helpful in showing pathologic tissue, of which it provided better or much better images in 73% of the cases. In 20% of the cases, tissue harmonic imaging yielded images of the same quality as fundamental scanning. In only 6% of the cases did tissue harmonic imaging of pathologic tissue yield images that were of worse or much worse quality (Fig. 3B). For pathologic tissue, tissue harmonic images were on average significantly better than fundamental images in all examination types (Table 2). In both normal and pathologic tissue, tissue harmonic imaging was found to be most helpful for obstetric, abdominal, superficial vascular, and musculoskeletal examinations and somewhat less helpful for pelvic and small-parts examinations.
|
|
Tissue harmonic images were more diagnostic than fundamental images in 43% of the cases. Tissue harmonic imaging was of the same diagnostic quality in half of the cases and of less diagnostic quality in only 8% of the cases (Fig. 3C). Table 3 shows the diagnostic value scores for tissue harmonic imaging using the 5-point scale. Only if tissue harmonic imaging was essential to the diagnosis did an examination score a 5 on this scale. On average, tissue harmonic imaging provided more diagnostic images than fundamental imaging for all except the pelvic examinations (Table 3). However, although the score for tissue harmonic imaging was not statistically different from that for fundamental imaging of the pelvis, tissue harmonic imaging did provide more diagnostic images in 38% (25/66) of the cases and was essential for the diagnosis in 5% (3/66) of the examinations. When all examination types are combined, tissue harmonic imaging was found to be essential for the diagnosis in 6% (11/200) of the cases.
|
|
Representative images comparing tissue harmonic imaging with fundamental imaging are shown in Figures 4A,4B,5A,5B,6A,6B,7A,7B.
|
|
|
|
|
|
|
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The tissue harmonic image has a somewhat different appearance from that generated by conventional fundamental pulse echo sonography. Tissue harmonic images have greater contrast. This can be advantageous in characterizing and evaluating fluid-containing areas, but it also requires learning new diagnostic standards for the normal appearance of solid structures. A single button activates tissue harmonic imaging, but it is important for the operator to optimize the image to avoid potential loss of information by a decrease in gray scale. The harmonic image is typically improved by using a higher near gain, a similar to lower far gain and increased display dynamic range, when compared with the equivalent fundamental scan. Adjustment of the system gain is also often helpful to optimize the display. There is a minor but noticeable decrease in frame rate in tissue harmonic images. We have not found this a significant limitation to its use even in the study of the fetal heart.
We subjectively noted that tissue harmonic images have somewhat less penetration than those produced by fundamental sonography. We do not find this a significant limiting factor in deep abdominal imaging. The capacity to transmit with fundamental frequencies as low as 1.2 MHz allows tissue harmonic imaging to penetrate deep structures as well as or better than fundamental sonography in all patients. Penetration is more noticeably limited in high-frequency transvaginal imaging. A difficulty in penetrating the upper limits of the normal or enlarged uteri transvaginally resulted in our scoring of tissue harmonic imaging as somewhat less efficacious than fundamental sonography in the normal pelvis. Despite this limitation, we found that transvaginal transducer tissue harmonic imaging improved our evaluation of many adnexal masses (Fig. 4A,4B) and that transabdominal tissue harmonic imaging was highly efficacious in the evaluation of a wide variety of pelvic and obstetric abnormalities (Fig. 5A,5B). Tissue harmonic imaging penetration was not a problem in small-parts applications.
We found tissue harmonic imaging efficacious or essential in a wide range of studies. In abdominal and transabdominal pelvic studies the improvements mirror those noted for frequency-based harmonic imaging. The normal kidneys, pancreas, and aorta are better delineated, especially in heavy patients or in patients with scoliosis. Uterine fibroids are often more clearly delineated. The depiction of fetal anatomy and placental structure and location in heavy patients is significantly improved. Tissue harmonic imaging significantly improves detection and characterization of renal and hepatic masses, gallstones, subtle hydronephrosis, appendicitis, and deep abdominal and pelvic fluid collections, particularly in obese patients.
In small-parts and transvaginal imaging, the greatest improvement in visualizing normal structures is in the deep veins during extremity compression studies, in characterizing the endometrium, and in recognizing ovaries and very early pregnancies. There is often a major improvement in differentiating solid breast masses from normal breast lobules; characterizing solid, cystic, and complex breast masses; delineating thyroid masses; and evaluating centrally placed adnexal masses.
Side lobe and reverberation artifacts often produce diagnostic confusion in the evaluation of hypoechoic breast masses. We found that tissue harmonic imaging significantly improves our ability to differentiate complex cysts from solid masses throughout the breast and to more accurately characterize masses, especially simple cysts, deep in fatty breasts. There is a major reduction of internal echoes in such structures and increased posterior acoustic enhancement (Fig. 6A,6B). Some isoechoic solid masses may become more evident on tissue harmonic imaging, but shadowing from normal structures such as Cooper's ligaments is also increased. Fibroadenomas are well shown on tissue harmonic imaging, but the high surrounding contrast can make characterization of their thin surrounding echogenic pseudocapsule difficult. We have found a combination of fundamental and harmonic imaging best when studying solid breast masses.
Compression studies of peripheral veins in large or edematous limbs are often compromised by artifacts and beam defocusing. Tissue harmonic imaging significantly improves the capacity to visualize veins in these problem cases and also allows more complete evaluation of focal abnormalities in the veins (Fig. 7A,7B).
No attempt was made to compare fundamental and harmonic images at the same frequency. The operators were instructed to choose optimum frequencies for both the fundamental and harmonic scans. Early in the study, the operators usually used the harmonic equivalent of the initial fundamental frequency specified by the preset for the type of examination they were performing. Operators with greater experience tended to use similar or lower harmonic frequencies rather than the equivalent fundamental frequency to improve penetration in transvaginal and deep abdominal scans and in evaluations of large fatty breasts. Similar or higher harmonic frequencies than the corresponding fundamental were usually chosen to image smaller breasts and other superficial structures.
A study of this type is necessarily limited by the subjective nature of the evaluations we report. Our study design was potentially biased, because all evaluators were aware of which kind of imaging was performed. Nevertheless, the overall conclusions were similar across the spectrum of observers.
In summary, this study shows that phase inversion tissue harmonic imaging can significantly improve the visualization of both normal and pathologic tissues. Tissue harmonic imaging often provides more diagnostic images and was found to be essential for diagnosis in some examinations. Its selective application has major diagnostic utility in a wide variety of clinical applications.
Acknowledgments
We thank the following sonographers without whom this work would not have
been possible: Jenna Cox-Mosburg, Rebecca Gefeller, Kelly Menninger, Patricia
Orin, Karyn Probasco, and Candace Spalding.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  S.-Y. Chiou, F. Forsberg, T. B. Fox, and L. Needleman Comparing Differential Tissue Harmonic Imaging With Tissue Harmonic and Fundamental Gray Scale Imaging of the Liver J. Ultrasound Med., November 1, 2007; 26(11): 1557 - 1563. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Scott, H. Swan, E. Jurriaans, and M. Voss Sonographic Detection of the Normal Appendix: Does the Combination of Real-Time Spatial Compounding and Tissue Harmonic Imaging Make a Difference? Journal of Diagnostic Medical Sonography, September 1, 2004; 20(5): 326 - 334. [Abstract] [PDF]
|
|
|
|
 |
|
 K. Strobel, M. Zanetti, L. Nagy, and J. Hodler Suspected Rotator Cuff Lesions: Tissue Harmonic Imaging versus Conventional US of the Shoulder Radiology, January 1, 2004; 230(1): 243 - 249. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. O. Oktar, C. Yucel, H. Ozdemir, A. Uluturk, and S. Isik Comparison of Conventional Sonography, Real-Time Compound Sonography, Tissue Harmonic Sonography, and Tissue Harmonic Compound Sonography of Abdominal and Pelvic Lesions Am. J. Roentgenol., November 1, 2003; 181(5): 1341 - 1347. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. M. Sofka and R. S. Adler Sonographic Evaluation of Shoulder Arthroplasty Am. J. Roentgenol., April 1, 2003; 180(4): 1117 - 1120. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
