HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Choi, H. Articles by Ma, J. Search for Related Content PubMed PubMed Citation Articles by Choi, H. Articles by Ma, J. Social Bookmarking                      What's this? Hotlight (NEW!) What's Hotlight?

Use of Perfluorocarbon Compound in the Endorectal Coil to Improve MR Spectroscopy of the Prostate

¹ Department of Diagnostic Radiology, The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd., Box Unit 368, Houston, TX 77030.
² Department of Imaging Physics, The University of Texas M. D. Anderson Cancer Center, Houston, TX.

Received August 8, 2007; accepted after revision October 23, 2007.

 
Presented at the 2004 scientific assembly and annual meeting of the Radiological Society of North America, Chicago, IL.

Address correspondence to H. Choi (hchoi{at}mdanderson.org).

Abstract

Top Abstract Introduction Subjects and Methods Results Discussion References

 
OBJECTIVE. The purpose of this study was to evaluate the utility of perfluorocarbon (PFC) compound compared with air in the endorectal coil in improving the quality of conventional MR spectroscopy of the prostate.

SUBJECTS AND METHODS. A total of 62 consecutively registered patients were selected. MR spectroscopy of the prostate was performed with PFC in the endorectal coil for 34 of the patients and with air for 28. In the cases of 13 of the 28 patients, MR spectroscopy was repeated with a PFC-filled endorectal coil. The spectral line widths and the spectral separations from MR spectroscopy with an air-filled endorectal coil were compared with those obtained with a PFC-filled coil.

RESULTS. In all 62 patients, the mean line width values were reduced, from 13.3 ± 3.0 Hz with air to 7.3 ± 2.0 Hz with PFC (p = 0.0001, Student's t test). In 13 patients who underwent MR spectroscopy with air followed by MR spectroscopy with PFC, the mean line width values were reduced, from 14.8 ± 3.4 Hz with air to 7.0 ± 1.5 Hz with PFC (p = 0.0001, Student's t test). In 72 voxels analyzed for spectral separation, clear separations between the choline, polyamine, and creatine peaks were found in 57 voxels with PFC and four with air. Better splitting of the citrate peaks was found in 35 voxels with PFC and one with air.

CONCLUSION. The use of PFC compound in the endorectal coil is a safe and cost-effective way to consistently generate high-quality prostate MR spectroscopic results and to substantially improve quantitation of prostatic metabolites. These improvements should increase the diagnostic value of MR spectroscopy in the care of patients with prostate cancer.

Keywords: carcinoma • MR spectroscopy • perfluorocarbon • prostate

Introduction

Top Abstract Introduction Subjects and Methods Results Discussion References

 
Prostate carcinoma is the second most common cancer death a mong American men [1]. The dia gnosis of prostate carcinoma is made with endorectal biopsy, often after an abnormal result for serum prostate-specific antigen or abnormal findings at digital rectal examination during a routine checkup. Once the diagnosis is made, the main goal of initial staging of prostate carcinoma is to assess risk of recurrence or treatment failure and to determine the optimal treatment [2]. This initial staging evaluation is typically based on a combination of prostate-specific antigen level, digital rectal examination findings, and biopsy results (Gleason score) [2]. However, the sensitivity of this clinical information in the diagnosis and staging of prostate carcinoma is suboptimal [3, 4]. MRI has become increasingly important in the management of prostate cancer, primarily in initial staging, although its accuracy for tumor detection is only moderate [5–7].

MR spectroscopy of the prostate is used to measure major prostatic metabolites, such as choline, creatine, and citrate, and in combination with MRI has been shown to significantly improve detection of tumors in the peripheral zone, primarily by improving specificity [5–7]. MR spectroscopy also has been found to have reasonable accuracy in the detection of transitional zone tumors [6] and to be more accurate than endorectal biopsy in the apex [8].

The commercially available endorectal coil currently used for MR spectroscopy of the prostate is designed to be inflated with air once the coil is introduced into the rectum. One of the limitations of using an air-filled endorectal coil is deteriorating magnetic field homogeneity caused by mismatch of magnetic susceptibility at the air–tissue interface. Spectral quality deteriorates, and metabolite peaks (e.g., choline and creatine peaks) cannot be well separated or accurately quantified. One way to improve magnetic field homogeneity is to replace the air in the endorectal coil with a material, such as perfluorocarbon (PFC), that has magnetic susceptibility closely matching that of prostate tissue [9]. A liquid form of PFC in a pillow (Sat Pad, Alliance Pharmaceutical) has been successfully used to improve local magnetic field homogeneity and fat saturation in the air–tissue interface of the neck and to improve the quality of MR images of the neck and cervical spine [9]. PFC also has been introduced as a versatile contrast agent [10, 11], such as a negative gastrointestinal contrast agent for MRI (Imagent GI, Alliance Pharmaceutical) [10].

View larger version (12K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 1 —Graph shows spectral line widths for prostate MR spectroscopy with both air and perfluorocarbon (PFC) in endorectal coils (n = 13). Mean line width decreased from 14.8 to 7.0 Hz with PFC-filled endorectal coil (p < 0.0001, Student's t test).

We have explored the feasibility of using an endorectal coil filled with PFC to resolve the magnetic susceptibility mismatch that occurs when the coil is filled with air. The magnetic field inhomogeneity is characterized by the magnetic field distribution within a volume of interest. Under the assumption that the field distribution is random and gaussian, the spectral shape of the water resonance can be shown to be Lorentzian. In this case, the spectral line width (full width at half maximum of the water peak) can be used to characterize magnetic field inhomogeneity. We describe our experience using this approach and present the results of a comparison of spectral line widths in the evaluation of the quality of prostate MR spectroscopy in which air in the endorectal coil was replaced with PFC.

Subjects and Methods

Top Abstract Introduction Subjects and Methods Results Discussion References

 
A total of 62 consecutively registered patients referred for MRI and MR spectroscopy of the prostate in the period February–September 2004 were selected. The mean age of the patients was 66 years (range, 51–77 years). All patients had biopsy-proven prostate carcinoma. A liquid form of PFC (FC-77, Fluorinert, 3M) was used in the study. Prostate MRI and MR spectroscopy were performed with endorectal coils filled with PFC for 34 of 62 patients and with air for 28. In the cases of 13 patients who had air-filled endorectal MR spectroscopic results of poor quality, MR spectroscopy was repeated with a PFC-filled endorectal coil at the same session or within 1 week. This study complied with the HIPAA and was conducted with the approval of the institutional review board with a waiver of informed consent.

Imaging Technique
All data were acquired on a 1.5-T MRI unit (Signa LX, GE Healthcare). After routine high-resolution endorectal MRI, 3D MR spectroscopy was performed with Prose software (GE Healthcare) with spectral water and fat suppression. The imaging parameters for 3D MR spectroscopy were TR/TE, 1,000/130; number of signals averaged, 1; field of view, 11 x 5.5 x 5.5 cm; matrix size, 16 x 8 x 8. Six very selective spatial saturation bands were applied to minimize fat contamination from outside the Prose volume box. One very selective spatial saturation band was placed at each of the four corners of the largest axial image, and two of the bands were placed on a midline sagittal image. For the patients who underwent MR spectroscopy with both air and PFC at the same session, the air in the coil from the first MR spectroscopic procedure was withdrawn as completely as possible before PFC was introduced into the coil for collection of the second data set. Approximately 100 cm³ of air and 90 mL of PFC were used in the endorectal coils.

View larger version (4K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 2 —Graph shows spectral line widths for prostate MR spectroscopy with either air or perfluorocarbon (PFC) in endorectal coils (n = 62). Mean line width decreased from 13.3 to 7.3 Hz with PFC-filled endorectal coil (p < 0.0001, Student's t test).

Data Collection and Analysis
The optimized spectral line widths were recorded after shimming and preimaging calibration for each MR spectroscopic data acquisition for all patients. Bivariate analysis was performed with the values of the spectral line widths measured in hertz with the air- and PFC-filled endorectal coils. Mean values and SD of the spectral line widths measured with the air- and PFC-filled endorectal coils were calculated. Paired Student's t tests were performed to compare the values of the spectral line widths measured with the two imaging methods.

In five of the 13 patients for whom MR spectroscopic data were obtained with both air and PFC, an experienced radiologist manually counted the number of voxels that showed clear separation of choline, polyamine, and creatine peaks and the number of voxels that showed a clear split of the citrate peak. For this analysis, all voxels were selected from regions completely within the prostate gland with overlaid axial T2-weighted images for guidance. The separation between choline, polyamine, and creatine peaks and splitting of the citrate peak of the spectra from MR spectroscopy with air were compared with the spectra from MR spectroscopy with PFC for the corresponding voxels. Voxels corresponding to the same anatomic locations from the two sets of spectral data were carefully and manually matched by the radiologist. When all choline, polyamine, and creatine peaks were identified separately, separation was considered clear. When the split of the citrate peak was at least 25% of the total height of the citrate peak, the split was considered clear.

Results

Top Abstract Introduction Subjects and Methods Results Discussion References

 
In the 13 patients who underwent MR spectroscopy with both air-filled endorectal coils and PFC-filled coils, the mean spectral line width with air was 14.8 ± 3.4 Hz (range, 9–20 Hz). The mean spectral line width with PFC was 7.0 ± 1.5 Hz (range, 5–10 Hz). Spectral line width was significantly lower with PFC-filled endorectal coils (p = 0.0001, Student's t test).

In all 62 patients who underwent MR spectroscopy with either air- or PFC-filled endorectal coils, the mean spectral line width with air was 13.3 ± 3.0 Hz (range, 9–20 Hz). For the same patient group, the mean spectral line width with PFC was 7.3 ± 2.0 Hz (range, 5–13 Hz). The line width was significantly lower with PFC-filled endorectal coils (p = 0.0001, Student's t test) (Figs. 1 and 2, Table 1).

In all but two patients, the spectral line widths were equal to or less than 10 Hz when PFC was used. On the images of these two patients, who underwent MR spectroscopy with PFC-filled endorectal coils and had spectral line widths of 12 and 13 Hz, a small amount of rectal air was layered between the anterior surface of the coil and the prostate. One of these patients also had extensive hemorrhage throughout the peripheral zone (Figs. 3A and 3B).

View larger version (151K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 3A —Axial MR images of patients with line widths greater than 10 Hz on perfluorocarbon (PFC)-filled endorectal MR spectroscopy. 65-year-old man with clinical stage T2c and Gleason score 7 adenocarcinoma of prostate, and line width of 12 Hz. T2-weighted axial image shows small amount of rectal air in layer anterior to PFC-filled endorectal coil (arrowheads).

View larger version (165K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 3B —Axial MR images of patients with line widths greater than 10 Hz on perfluorocarbon (PFC)-filled endorectal MR spectroscopy. 64-year-old man with clinical staage T1c and Gleason score 6 adenocarcinoma of prostate, and line width of 13 Hz. T1-weighted axial image shows small amount of rectal air in layer anterior to PFC-filled endorectal coil (arrowheads). Diffuse hemorrhage is present throughout peripheral zone (arrow).

View larger version (144K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 4A —50-year-old man with clinical stage T1c and Gleason score 7 adenocarcinoma of prostate and improvement in spectral resolution of MR spectroscopy with use of perfluorocarbon (PFC)-filled endorectal coil. T2-weighted axial image shows benign-appearing hyperintense voxel (box) in left peripheral zone (arrows).

View larger version (11K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 4B —50-year-old man with clinical stage T1c and Gleason score 7 adenocarcinoma of prostate and improvement in spectral resolution of MR spectroscopy with use of perfluorocarbon (PFC)-filled endorectal coil. Graph shows nondiagnostic quality of MR spectra from voxel in A obtained with air-filled endorectal coil.

View larger version (13K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 4C —50-year-old man with clinical stage T1c and Gleason score 7 adenocarcinoma of prostate and improvement in spectral resolution of MR spectroscopy with use of perfluorocarbon (PFC)-filled endorectal coil. Graph shows diagnostic MR spectra from repeated MR spectroscopy with PFC-filled endorectal coil clearly resolve choline and creatine peaks and additional polyamine peak (Po) between them. Line width decreased from 12 Hz with air to 6 Hz with PFC in endorectal coils.

View larger version (133K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 5A —64-year-old man with clinical stage T1c adenocarcinoma of prostate with Gleason score of 6 and improved spectral resolution of MR spectroscopy with perfluorocarbon (PFC)-filled endorectal coil. MR spectra from two hypointense voxels on T2-weighted axial image (A) performed with air (B) and PFC (C) in endorectal coils show dramatic improvement in spectral resolution with PFC. Polyamine peaks (Po) are clear in C but not in B. Line width decreased from 16 to 6 Hz. Larger MR spectroscopy volume box for MR spectroscopy with PFC (white, A) than for MR spectroscopy with air (black, A) allows inclusion of larger area of prostate tissue and anterior rectum.

View larger version (13K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 5B —64-year-old man with clinical stage T1c adenocarcinoma of prostate with Gleason score of 6 and improved spectral resolution of MR spectroscopy with perfluorocarbon (PFC)-filled endorectal coil. MR spectra from two hypointense voxels on T2-weighted axial image (A) performed with air (B) and PFC (C) in endorectal coils show dramatic improvement in spectral resolution with PFC. Polyamine peaks (Po) are clear in C but not in B. Line width decreased from 16 to 6 Hz. Larger MR spectroscopy volume box for MR spectroscopy with PFC (white, A) than for MR spectroscopy with air (black, A) allows inclusion of larger area of prostate tissue and anterior rectum.

View larger version (14K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 5C —64-year-old man with clinical stage T1c adenocarcinoma of prostate with Gleason score of 6 and improved spectral resolution of MR spectroscopy with perfluorocarbon (PFC)-filled endorectal coil. MR spectra from two hypointense voxels on T2-weighted axial image (A) performed with air (B) and PFC (C) in endorectal coils show dramatic improvement in spectral resolution with PFC. Polyamine peaks (Po) are clear in C but not in B. Line width decreased from 16 to 6 Hz. Larger MR spectroscopy volume box for MR spectroscopy with PFC (white, A) than for MR spectroscopy with air (black, A) allows inclusion of larger area of prostate tissue and anterior rectum.

Discussion

Top Abstract Introduction Subjects and Methods Results Discussion References

 
The commercially available endorectal coil for prostate MR spectroscopy is designed to be inflated with air within the rectum for better coil positioning and improved coil coverage. The ballooned coil also supports the prostate and minimizes motion. Despite technical improvements in prostate MR spectro scopy with an air-filled endorectal coil, clear separation of the prostatic metabolites of in terest, such as choline and creatine, is not readily achievable in many voxels. In a pre liminary analysis of poor-quality MR spectroscopy data at our institution (Garcia M, et al., presented at the 2006 annual meeting of the Society of Magnetic Resonance Technologists), we identified the main under lying cause as magnetic field hetero geneity inherent in the interface between the air-filled coil and the rectum.

Utility of PFC in improving magnetic susceptibility in MR spectroscopy of the prostate was found in a phantom study [13] and in a study with two independent groups of prostate cancer patients [14]. Recently, Prando et al. [15] showed improvement in spectral quality by using PFC in a prospective study investigating patients with elevated prostate-specific antigen and prior negative biopsy. In our study, we quantitatively compared, within and be tween patients, the quality of MR spectro scopy performed with air with that of MR spectroscopy performed with PFC in the endorectal coils. Our findings confirmed that the use of PFC within the coil significantly improves spectral line width by improving the susceptibility match between the PFC and the prostate tissue (p = 0.0001, Student's t test) (Table 1). Con sequently, we found that better separation, not only between the choline and creatine peaks but also between the polyamine and creatine peaks, was more readily achievable and that spectral im provement with PFC often changed a non diagnostic spectroscopic examination with air into a diag nostic one (Figs. 4A, 4B, 4C, 5A, 5B, and 5C). In our limited study population, we also found sub stantial improvement in the actual spectral separation between the choline, polyamine, and creatine peaks and in the quality of the citrate peak (Table 2). The improvement in spectral separation in the general patient population may be lower than reported in this study because we analyzed only nondiagnostic MR spectroscopy data obtained with air.

Our study clearly showed that identification of the additional polyamine peak was possible with the introduction of PFC. Along with citrate, polyamine is found at a sub stantially high concentration in the epithelial cells of the prostate gland and is known to have important roles in proliferation, differ entiation, and growth of glandular epithelial cells. Over expression of ornithine decarbo xylase, a polyamine biosynthesis enzyme, has been found in prostate epithelial cells and a strong reduction of polyamine in prostate cancer cells [16, 17]. Thus detection of changes in the polyamine peak may play an important role in the treatment of patients with prostate cancer.

Even when a small amount of rectal air was found between the PFC-filled endorectal coil and the prostate, we found that spectral line width was better than the mean line width measured when an air-filled endorectal coil was used. Accumulation of this small amount of rectal air owing to improper contact between the coil and the anterior wall of the rectum probably occurred because the PFC-filled endorectal coil is heavier than the air-filled endorectal coil. To maximize the benefit of PFC, careful observation is required to identify air between the prostate and the coil and to adjust the coil position before the start of data acquisition.

An additional benefit of using PFC is that a greater amount of prostatic tissue can be included in the spectroscopic volume. The ballooned coil with air often deforms the prostate, wrapping around the anterior surface of the rectum. It was often necessary to exclude the posterolateral portion of the peripheral zone to avoid the air within the spectroscopic volume and to maximize field homogeneity. Using PFC instead of air in the endorectal coil increases latitude in selection of the spectroscopic volume (Figs. 5A, 5B, and 5C).

The FC-77 used in the study is a chemically inert clear liquid without odor and virtually nontoxic to the human body (FC-77 Material Safety Data Sheet, 3M). At the time of our study, the cost of PFC per procedure was deemed inconsequential (less than 2% of the cost of the endorectal coil). FC-77 has a practically indefinite shelf life and can be easily and safely disposed of as needed. At our institution, it is collected by the depart ment of environmental safety in a plastic container to be shipped back to the manu facturing company. It should be noted that FC-77 has not yet been approved by the U.S. Food and Drug Administration (FDA) for commercial use in the endorectal coil. Nonetheless, the FDA has designated FC-77 a nonsignificant risk medical device (Office of Compliance, The M. D. Anderson Cancer Center; written communication, April 2005) and permits its use when appropriate guide lines are followed [17]. A commercial 3-T endorectal coil has been approved by the FDA for use in conjunction with a PFC-based product.

We expect that the improvement in field heterogeneity observed with the PFC-filled endorectal coil setup will be applicable to prostate MR spectroscopy with higher-field-strength systems, such as 3-T units. Without improvement in field inhomogeneity, spec tral broadening will be proportional to field strength, negating the benefits of expected larger spectral separation at the higher field strength.

The results of our study indicate that use of PFC is safe and cost-effective for consistent generation of high-quality prostate MR spectra at 1.5 T. Use of this agent substantially improves quantitation of prostatic metabolites and has the advantages of an air-filled coil setup. These improvements are expected to increase the value of MR spectroscopy of the prostate in the clinical management of prostate cancer by improving the specificity of tumor detection within the gland.

Acknowledgments
 
We thank Donald Podoloff, Division of Diagnostic Imaging, The University of Texas M. D. Anderson Cancer Center, for administrative support; Joe Zhou, University of Chicago, for early technical support; and Michelle Garcia for technical assistance throughout the study.

References

Top Abstract Introduction Subjects and Methods Results Discussion References

 

1. American Cancer Society. Cancer Facts and Figures 2007. Atlanta, GA: American Cancer Society, 2007. http://www.cancer.org/downloads/STT/CAFF2007PWSecured.pdf. Accessed January 8, 2008
2. Jacobson JO, Brodsky G, DuBeau C, et al. Cancer of the prostate. In: Osteen RT, ed. Cancer manual. Boston, MA: American Cancer Society, 1996:446 -461
3. Arcangeli CG, Ornstein DK, Keetch DW, Andriole GL. Prostate-specific antigen as a screening test for prostate cancer: the United States experience. Urol Clin North Am1997; 24:299 -306[CrossRef][Medline]
4. Whelan P. Multidisciplinary symposium: prostate cancer—planning primary therapy. Cancer Imaging2000; 1:44 -51[Medline]
5. Casciani E, Polettini E, Bertini L, et al. Contribution of the MR spectroscopic imaging in the diagnosis of prostate cancer in the peripheral zone. Abdom Imaging 2007; Feb 10 [Epub ahead of print]
6. Futterer JJ, Scheenen TW, Heijmink SW, et al. Standardized threshold approach using three-dimensional proton magnetic resonance spectroscopic imaging in prostate cancer localization of the entire prostate. Invest Radiol 2007;42 : 116-122[CrossRef][Medline]
7. Scheidler J, Hricak H, Vigneron DB, et al. Prostate cancer: localization with three-dimensional proton MR spectroscopic imaging: clinicopathologic study. Radiology 1999;213 : 473-480[Abstract/Free Full Text]
8. Wefer AE, Hricak H, Vigneron DB, et al. Sextant localization of prostate cancer: comparison of sextant biopsy, magnetic resonance imaging and magnetic resonance spectroscopic imaging with step section histology. J Urol 2000; 164:400 -404[CrossRef][Medline]
9. Eilenberg SS, Tartar VM, Mattrey RF. Reducing magnetic susceptibility differences using liquid fluorocarbon pads (Sat Pad): results with spectral presaturation of fat. Artif Cells Blood Substit Immobil Biotechnol 1994; 22:1477 -1483[CrossRef][Medline]
10. Brown JJ. Gastrointestinal contrast agents for MR imaging. Magn Reson Imaging Clin N Am 1996;4 : 25-35[Medline]
11. Mattrey RF. The potential role of perfluorochemicals (PFCs) in diagnostic imaging. Artif Cells Blood Substit Immobil Biotechnol 1994; 22:295 -313[Medline]
12. F2 Chemicals Limited. For selective direct fluorination and Flutec perfluorocarbon fluids: medical applications. http://www.fluoros.co.uk/data/medical_applications.php. Accessed January 8, 2008
13. Choi H, Zhou X, Ma J. Use of perfluorocarbon (PFC) in magnetic resonance spectroscopy (MRS) of prostate: a method to improve the linewidth and quality of spectra (abstr SSA07-02). RSNA 2004. Oak Brook, IL: Radiological Society of North America, 2004. http://archive.rsna.org/index.cfm?em_id=4407696&event=1. Accessed January 8, 2008
14. Guclu CC, Figueiredo E, Prando A, Racy D, Becerra R. A method to improve the spectral resolution in prostate spectroscopy (abstract). In: 2004 Scientific Meeting proceedings. Berkeley, CA: International Society for Magnetic Resonance in Medicine, 2004:11
15. Prando A, Kurhanewicz J, Borges AP, Oliveira EM Jr, Figueiredo E. Prostate biopsy directed with endorectal MR spectroscopic imaging findings in patients with elevated prostate specific antigen levels and prior negative biopsy findings: early experience. Radiology2005; 236:903 -910[Abstract/Free Full Text]
16. Romijn JC. Molecular and cellular biology of prostate cancer. In: Karr JP, Coffey DS, Smith RG, Tindall DJ, eds. Polyamine requirement of prostate cancer cell proliferation. New York, NY: Plenum, 1991: 103-114
17. Schipper RG, Romijn JC, Cuijpers VM, Verhofstad AA. Polyamines and prostatic cancer. Biochem Soc Trans 2003;31 : 375-380[CrossRef][Medline]
18. Food and Drug Administration. Information sheet guidance for IRBs, clinical investigators, and sponsors: significant risk and nonsignificant risk medical device studies. Rockville, MD: Food and Drug Administration, 2006. http://www.fda.gov/oc/ohrt/irbs/devrisk.pdf. Accessed January 8, 2008

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This Article Abstract Figures Only Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Choi, H. Articles by Ma, J. Search for Related Content PubMed PubMed Citation Articles by Choi, H. Articles by Ma, J. Social Bookmarking                      What's this? Hotlight (NEW!) What's Hotlight?