AJR ARRS Member Benefits
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Melhem, E. R.
Right arrow Articles by Itoh, R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Melhem, E. R.
Right arrow Articles by Itoh, R.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?
AJR 2003; 180:65-69
© American Roentgen Ray Society

Defining Thresholds for Changes in Size of Simulated T2-Hyperintense Brain Lesions on the Basis of Qualitative Comparisons

Elias R. Melhem1,2,3, Edward H. Herskovits1,4, Kader Karli-Oguz1, Xavier Golay3, Dima A. Hammoud1, Brian J. Fortman1, Fletcher M. Munter1 and Ryuta Itoh1,3

1 Division of Neuroradiology, The Johns Hopkins Medical Institutions, 600 N. Wolfe St., Baltimore, MD 21287-7619.
2 MR Perception Laboratory and the F. M. Kirby Center for Functional Magnetic Resonance Imaging of The Kennedy Krieger Institute, Baltimore, MD 21287-7619.
3 Present address: Department of Radiology, Hospital of the University of Pennsylvania, 3400 Spruce St., Philadelphia, PA 19104.
4 Department of Biostatistics, The Johns Hopkins School of Public Health, Baltimore, MD 21287-7619.

Received April 9, 2002; accepted after revision July 11, 2002.

 
Address correspondence to E. R. Melhem.

Supported in part by National Institutes of Health grant R01 AG13743.


Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
OBJECTIVE. Our purpose was to define thresholds below which trained reviewers cannot detect changes in the size of T2-hyperintense brain lesions.

MATERIALS AND METHODS. We generated T2-weighted brain MR images (TR/TE, 4000/80) with simulated hyperintense lesions derived from a real multiple sclerosis plaque. The size of the original multiple sclerosis lesion was varied by scaling up or down the lesion using a bicubic interpolation method. Three hundred seventy-eight composite images, in which two T2-weighted images containing lesions were paired, were presented to three equally trained neuroradiologists to define thresholds below which changes in original lesion size could not be detected. Stepwise logistic regression was used to evaluate the dependency of size thresholds on the original size of the lesion.

RESULTS. Thresholds ranged from a 5% to 15% increase in the original lesion diameter. For increases greater than 15%, all three reviewers detected the change in lesion size irrespective of the diameter of the original lesion. There was a dependency of the threshold on the diameter of the original lesion (p = 0.02).

CONCLUSION. Using an MR simulator, we can define thresholds below which changes in original lesion size cannot be reliably detected. These results may guide the design of clinical trials that rely on trained reviewers to assess change in lesion burden.


Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Change in T2-hyperintense brain lesion load assessed on MR imaging is a widely used secondary end point in clinical studies involving conditions such as multiple sclerosis, sickle cell disease, closed head injury, and the aging brain [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18]. However, poor correlations between change in brain lesion load and primary clinical end points such as neurologic disability scores, especially in multiple sclerosis, have been attributed in part to inaccuracies in the measurement of load [7, 8]. Other causes are subjectivity and the nonlinearity of primary clinical end points and the inability of conventional MR imaging to show the true lesion load and to reliably differentiate pathologic entities that result in T2-hyperintense lesions such as edema, inflammation, demyelination, microcystic, and macrocystic leukomalacia [7, 8].

Assessment of change in brain lesion load can be achieved using qualitative (visual inspection by trained reviewers) [16, 18] or quantitative [7, 8, 19, 20] methods. Quantitative methods can be further divided into fully automated methods and semiautomated methods that rely on manual outlining of lesions by trained reviewers followed by computer-based quantification [7]. However, the fully automated methods continue to require validation against the current standard of manual outlining by trained reviewers [7].

Using simulated T2-hyperintense brain lesions placed in computer-generated T2-weighted brain MR images, we sought to define thresholds below which trained reviewers cannot detect changes in lesion size. We hypothesized that size thresholds do exist, that size thresholds are dependent on the original size of the lesion, and that size thresholds vary with the reviewer.


Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
MR Data Acquisition
MR imaging was performed on a 1.5-T MR system (ACS-NT; Philips Medical Systems, Best, The Netherlands) with a maximal gradient capability of 23 mT · m-1 and a slew rate of 103 mT · m-1 · msec-1 using a quadrature head coil operating in receive mode.

A mixed multiecho, spin-echo, and inversion recovery MR sequence was used to obtain images of the brain of a 40-year-old consenting male volunteer at the level of the lateral ventricles. The mixed sequence simultaneously provided two image data sets: eight spin-echo images (TR, 1500; 1 signal acquired) with different TEs (20, 40, 60, 80, 100, 120, 140, 160) and eight inversion recovery images (TR, 2000; inversion time, 400 msec; 1 signal acquired) at the same TEs. Image slice thickness was 5 mm, inplane resolution was 0.80 x 0.86 mm (rectangular field of view, 165 x 220 mm; scan matrix, 205 x 256), and acquisition time was 9 min 30 sec.

Brain-Image Simulation
The generation of normal brain images involved two steps: First, we generated pixel-by-pixel T1-relaxation, T2-relaxation, and proton density ({rho}) brain maps (256 x 256 matrix) at the level of the lateral ventricles online (software release 6.2, Philips Medical Systems) using the image data sets from the mixed MR sequence. Second, we generated images simulating T2-weighted MR sequences offline (Enterprise 5500; Sun Microsystems, Mountain View, CA) using T1-relaxation, T2-relaxation, and proton density pixel values from the corresponding maps.

We developed an image-simulation software using Interactive Data Language (Research Systems, Boulder, CO). Each pixel value S(x, y) of the generated image was calculated using the following equation:

We derived T1(x, y), T2(x, y), and {rho}(x, y) from the corresponding pixel values of the T1-relaxation, T2-relaxation, and proton density maps, respectively. We selected parameters for the T2-weighted MR sequence on the basis of values typically used in clinical brain imaging: TR/TE, 4000/80.

Lesion Simulation
Using the same mixed multiecho, spin-echo, and inversion recovery MR sequence described previously, we obtained images of the brain of a 32-year-old consenting woman with a known multiple sclerosis lesion in the left centrum semiovale. We derived T1(x, y), T2(x, y), and {rho}(x, y) for all the pixels in the multiple sclerosis lesion from the corresponding pixel values of the T1-relaxation, T2-relaxation, and proton density maps, respectively (Fig. 1). Only pixels with T1, T2, and {rho} values two standard deviations or more above the average pixel values in the corresponding normal-appearing white matter (right centrum semiovale) were included in the multiple sclerosis lesion.



View larger version (13K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 1. —Color-coded T1, T2, and proton density ({rho}) pixel maps of multiple sclerosis lesion showing variability in pixel values in lesion.

 

Nine simulated original lesions with different diameters ranging from 4 to 12 mm by increments of 1 mm were generated on the basis of the pixel values of the multiple sclerosis lesion. We enlarged original lesions 20 times by increments of 5%, resulting in a group of 21 corresponding lesions in which their diameters were equal to: X + X · I · 0.05 (X is the diameter of the original lesion, and I is an integer ranging from 0 to 20 by increments of 1) (Fig. 2A,2B,2C,2D,2E,2F).



View larger version (116K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 2A. —Magnified view of simulated images containing examples of simulated lesions placed in right parietooccipital lobe. Simulated T2-weighted MR image (TR/TE, 4000/80) shows original lesion (6 mm).

 


View larger version (110K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 2B. —Magnified view of simulated images containing examples of simulated lesions placed in right parietooccipital lobe. Simulated T2-weighted MR images (4000/80) show lesions 10% larger than original lesion (6.6 mm) (B)

 


View larger version (109K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 2C. —Magnified view of simulated images containing examples of simulated lesions placed in right parietooccipital lobe. Simulated T2-weighted MR images (4000/80) show lesions 20% larger than original lesion (7.2 mm) (C)

 


View larger version (111K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 2D. —Magnified view of simulated images containing examples of simulated lesions placed in right parietooccipital lobe. Simulated T2-weighted MR images (4000/80) show lesions 30% larger than original lesion (7.8 mm) (D)

 


View larger version (112K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 2E. —Magnified view of simulated images containing examples of simulated lesions placed in right parietooccipital lobe. Simulated T2-weighted MR images (4000/80) show lesions 40% larger than original lesion (8.4 mm) (E)

 


View larger version (113K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 2F. —Magnified view of simulated images containing examples of simulated lesions placed in right parietooccipital lobe. Simulated T2-weighted MR images (4000/80) show lesions 50% larger than original lesion (9.0 mm) (F)

 

We varied the size of the original multiple sclerosis lesion by scaling up or down the lesion using a bicubic interpolation method, allowing for a subsampling scaling by steps of 0.20 mm. The interpolated contour of the lesion was determined by embedding the lesion in the brain using a 5% threshold. Each of the simulated lesions was placed in the same location (white matter of the right parietooccipital region) on a separate T2-weighted MR image.

For the sake of comparison, composite images in which T2-weighted MR images containing lesions of one group were paired side by side with the image containing the original lesion of that group. The image containing the original lesion was placed on the left of the composite image 20 times and on the right 20 times. In addition, two extra composite images, in which two identical images containing the original lesion were placed side by side, were generated. This process resulted in a total of 42 composite images for each of the nine groups.

Image Presentation
Three neuroradiologists of equivalent training (instructor level) evaluated each of the 378 composite images in the same randomized order. None of the reviewers were involved in any other aspect of the study. The reviewers received instructions that included the aim of the study, presentation of sample images with simulated lesions, and a 5-min practice session immediately preceding the reviewing session. They were also informed that a total of 378 composite images would be shown to them in one session, that images would be presented in random order, and that the reviewing time for one image should be minimized and must not exceed 20 sec.

Reviewers examined one composite image at a time on a 21-inch (53 cm) monitor (viewable composite image size, 16 x 12 inches [40.64 x 30.48 cm], 1600 dots [horizontal] x 1200 lines [vertical]). The window width and level were adjusted across images, using an Interactive Data Language routine based on the equation:

where datamin is minimum signal intensity and datamax is maximum signal intensity, to match values used in clinical brain imaging. An experienced neuroradiologist confirmed the similarity between the appearances of the generated images and those of actual clinical images.

Reviewers scored each of the composite images on the basis of a 3-point scoring system: 0, simulated lesion in image on left is equal in size to lesion in image on right; 1, simulated lesion in image on left is larger than lesion in image on the right; and 2, simulated lesion in image on left is smaller than lesion in image on right.

One reviewer reevaluated each of the 378 composite images for assessment of intrarater reliability after a 7-week interval, as previously described.

Data Analysis
Thresholds below which at least one reviewer assigned the incorrect score (false-negative and false-positive reviews) were identified.

Stepwise logistic regression was used to evaluate the dependency of size thresholds on the original size of the lesion, on the side (whether in the composite image, the image containing the original lesion was placed on the left or on the right), and on the reviewer. The variables used for the model were the reviewers' correct or incorrect responses (binary independent variable), the diameter of the original lesion (continuous dependent variable), the side (binary dependent variable), and the reviewer (categoric dependent variable). A p value of less than 0.05 was considered statistically significant.

The kappa statistic was used for pairwise evaluation of inter- and intrarater reliability.


Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Each reviewer examined 378 composites. The total time for a reviewing session ranged from 1 to 1.5 hr per reviewer.

Thresholds ranged from 5% to 15% increase in the original lesion diameter. For increases greater than 15%, all three reviewers assigned the correct score irrespective of the diameter of the original lesion.

Stepwise logistic regression showed a dependency of the threshold on the diameter of the original lesion (p = 0.02) with lower thresholds found at a larger original lesion diameter (Fig. 3). A weak but significant difference in the threshold was also found among the three reviewers (p = 0.03). No difference was shown based on the side (p = 0.85).



View larger version (8K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		Fig. 3. —Plot of thresholds below which at least one reviewer assigned incorrect score compared to original lesion size. Plot shows that trained reviewers can detect smaller percentage increases in diameter of larger original lesion size.

 

Because of complete agreement among the three reviewers for increases beyond 15%, the kappa statistic for evaluation of inter- and intrarater reliability was performed only for lesions with diameters that were 0-15% larger than the original lesions. For interrater reliability, there was excellent agreement between reviewers 1 and 2 ({kappa} = 0.77) and reviewers 2 and 3 ({kappa} = 0.73) and good agreement between reviewers 1 and 3 ({kappa} = 0.56). For intrarater reliability, there was excellent review—rereview agreement ({kappa} = 0.92).


Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
The advantages of MR imaging over primary clinical end points for evaluation of brain lesion load include providing more objective and continuous measures amenable to established statistical analyses, allowing more rapid assessment of the effects of therapy, and having less chance for reviewer unblinding [7, 8]. However, the reproducibility and accuracy of MR imaging—based quantification of lesion load can be influenced by factors related to MR imaging hardware and software (such as magnetic field strength, stability, and homogeneity; gradient performance; and pulse sequence parameters), patient repositioning, and segmentation techniques [21,22,23,24,25,26,27].

Qualitative assessment (visual inspection by trained reviewers) and manual outlining of brain lesions remain integral steps for lesion load assessment in clinical trials [7, 8]. Our results show that trained reviewers can reliably detect a 15% or more increase in diameter for all original lesion sizes and a 10% or more increase in diameter for original lesions greater than 6 mm. For increases below these thresholds, there is good to excellent agreement among reviewers. This information may guide the design of clinical trials that rely on trained reviewers to assess changes in lesion burden on MR images. For example, investigators should be aware that the effectiveness of a specific drug designed to stop the progression of brain disease can only be questioned when the increase in lesion diameter exceeds the defined threshold. Our results also show that despite good to excellent agreement among reviewers (kappa statistic), the stepwise logistic regression showed a weak but significant difference in the threshold among the three reviewers (p = 0.03). This finding should caution investigators not to rely simply on the level of training when choosing reviewers for their trials but to test the interrater reliability of reviewers before recruitment using tools such as an MR simulator.

An interesting observation made from this study is the ability of the trained reviewers to detect increases in lesion diameter below the spatial resolution of the image. For example, a 10% increase in the diameter of a 6- or 7-mm lesion is below the in-plane spatial resolution of the image (0.80 x 0.86 mm). Although not addressed by the study design, this observation is probably due to the ability of trained reviewers to interpret increases in edge-pixel signal intensity as an increase in lesion size [28]. The bicubic interpolation method used in this study assigns greater signal intensity in the edge pixels for subpixel increases in diameter and defines a lower limit of 0.20 mm for detectable change. This new lower limit imposed by the interpolation method precludes us from studying original lesions with diameters of less than 4 mm.

Unlike previous studies that rely on phantoms [29], our study aimed to introduce an MR simulator containing more realistic-appearing simulated lesions based on a real multiple sclerosis lesion with the flexibility of assigning individual lesion size. Using this simulator, we could evaluate the effect of lesion number and location and the type of MR pulse sequence (T2-weighted vs fluid-attenuated inversion recovery vs proton density—weighted) on reviewers' sensitivity, specificity, and reliability [30]. In the future, this simulator may help assess the reliability of a reviewer's outlining the lesions before and during a clinical trial that uses brain imaging as a surrogate marker. We also envision this simulator being used to determine accuracy and reliability as well as to establish minimal performance standards for fully automated computer-based methods designed to quantify lesion load.

Our method for calculating T1, T2, and proton density maps from image data generated by a multiecho spin-echo sequence interleaved with a multiecho inversion recovery MR sequence, is based on the ratios and least squares algorithm [31]. This method has been validated with phantoms and human volunteers, and the calculated T1, T2, and proton density values are in agreement with values obtained using spectrometry [31].

Potential limitations of the simulator include the use of an approximate solution to the Bloch equation to generate the weighted images. A more exact solution includes a denominator (1 + e-(TR/T1) e-(TR/T2)) [32, 33]. The approximate solution is used in our simulations because the effect of the denominator on the signal intensity from brain parenchyma (gray and white matter) and simulated lesions is less than 1% on the T2-weighted MR images. Furthermore, using the approximate solution simplifies the calculation of the images and simulated lesions.

Because our simulation was based on conventional spin-echo sequences, we were unable to study the effects of fast imaging techniques, such as rapid acquisition with relaxation enhancement or echoplanar readouts, on the point-spread function of the simulated lesions and hence on the ability of trained reviewers to detect changes in lesion size [24, 34,35,36,37].

In conclusion, using an MR simulator containing simulated lesions based on a real multiple sclerosis lesion, we could define thresholds below which trained reviewers could not detect changes in lesion size. We also showed that these size thresholds are dependent on the original size of the lesion and vary among equally trained reviewers. Our results may guide the design of clinical trials that rely on trained reviewers to assess change in lesion burden on MR images.


References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Miller DH. Magnetic resonance in monitoring the treatment of multiple sclerosis. Ann Neurol 1994;36[suppl]:S91 -S94
  2. Miller DH, Albert PS, Barkhof F, et al. Guidelines for the use of magnetic resonance techniques in monitoring the treatment of multiple sclerosis: US National MS Society Task Force. Ann Neurol 1996;39:6 -16[Medline]
  3. Carswell H. MS researchers assess MRI as outcome tool. Diagn Imaging 1996;18:49 -51
  4. Zhao GJ, Li DK, Wolinsky JS, et al. Clinical and magnetic resonance imaging changes correlate in a clinical trial monitoring cyclosporine therapy for multiple sclerosis: The MS Study Group. J Neuroimaging 1997;7:1 -7[Medline]
  5. D'yachkova Y, Petkau J, White R. Longitudinal analyses for magnetic resonance imaging outcomes in multiple sclerosis clinical trials. J Biopharm Stat 1997;7:501 -531[Medline]
  6. Erickson BJ, Noseworthy JH. Value of magnetic resonance imaging in assessing efficacy in clinical trials of multiple sclerosis therapies. Mayo Clin Proc 1997;72:1080 -1089[Medline]
  7. Evans AC, Frank JA, Antel J, Miller DH. The role of MRI in clinical trials of multiple sclerosis: comparison of image processing techniques. Ann Neurol 1997;41:125 -132[Medline]
  8. Filippi M, Horsfield MA, Ader HJ, et al. Guidelines for using quantitative measures of brain magnetic resonance imaging abnormalities in monitoring the treatment of multiple sclerosis. Ann Neurol 1998;43:499 -506[Medline]
  9. Castelijns JA, Barkhof F. Magnetic resonance (MR) imaging as a marker for multiple sclerosis. Biomed Pharmacother 1999;53:351 -357[Medline]
  10. Stevenson VL, Miller DH. Magnetic resonance imaging in the monitoring of disease progression in multiple sclerosis. Mult Scler 1999;5:268 -272[Abstract/Free Full Text]
  11. Molyneux PD, Miller DH, Filippi M, et al. The use of magnetic resonance imaging in multiple sclerosis treatment trials: power calculations for annual lesion load measurement. J Neurol 2000;247:34 -40[Medline]
  12. Verlhac S, Bernaudin F, Tortrat D, et al. Detection of cerebrovascular disease in patients with sickle cell disease using transcranial Doppler sonography: correlation with MRI, MRA and conventional angiography. Pediatr Radiol 1995;25[suppl]:S14 -S19
  13. Adams RJ. Stroke prevention and treatment in sickle cell disease. Arch Neurol 2001;58:565 -568[Free Full Text]
  14. Brown RT, Davis PC, Lambert R, Hsu L, Hopkins K, Eckman J. Neurocognitive functioning and magnetic resonance imaging in children with sickle cell disease. J Pediatr Psychol 2000;25:503 -513[Abstract/Free Full Text]
  15. Levin HS, Williams DH, Eisenberg HM, High WM Jr, Guinto FC Jr. Serial MRI and neurobehavioral findings after mild to moderate closed head injury. J Neurol Neurosurg Psychiatry 1992;55:255 -262[Abstract]
  16. Bryan RN, Wells SW, Miller TJ, et al. Infarctlike lesions in the brain: prevalence and anatomic characteristics at MR imaging of the elderly—data from the Cardiovascular Health Study. Radiology 1997;202:47 -54[Abstract/Free Full Text]
  17. Manolio TA, Kronmal RA, Burke GL, et al. Magnetic resonance abnormalities and cardiovascular disease in older adults: the Cardiovascular Health Study. Stroke 1994;25:318 -327[Abstract]
  18. Bryan RN, Manolio TA, Schertz LD, et al. A method for using MR to evaluate the effects of cardiovascular disease on the brain: the Cardiovascular Health Study. AJNR 1994;15:1625 -1633[Abstract]
  19. Kikinis R, Shenton ME, Gerig G, et al. Routine quantitative analysis of brain and cerebrospinal fluid spaces with MR imaging. J Magn Reson Imaging 1992;2:619 -629[Medline]
  20. Mitchell JR, Karlik SJ, Lee DH, Eliasziw M, Rice GP, Fenster A. The variability of manual and computer assisted quantification of multiple sclerosis lesion volumes. Med Phys 1996;23:85 -97[Medline]
  21. Filippi M, van Waesberghe JH, Horsfield MA, et al. Interscanner variation in brain MRI lesion load measurements in MS: implications for clinical trials. Neurology 1997;49:371 -377[Abstract/Free Full Text]
  22. Filippi M, Marciano N, Capra R, et al. The effect of imprecise repositioning on lesion volume measurements in patients with multiple sclerosis. Neurology 1997;49:274 -276[Abstract/Free Full Text]
  23. Rovaris M, Gawne-Cain ML, Sormani MP, Miller DH, Filippi M. The effect of repositioning on brain MRI lesion load assessment in multiple sclerosis: reliability of subjective quality criteria. J Neurol 1998;245:273 -275[Medline]
  24. Rovaris M, Gawne-Cain ML, Wang L, Miller DH. A comparison of conventional and fast spin-echo sequences for the measurement of lesion load in multiple sclerosis using a semi-automated contour technique. Neuroradiology 1997;39:161 -165[Medline]
  25. Udupa JK, Wei L, Samarasekera S, Miki Y, van Buchem MA, Grossman RI. Multiple sclerosis lesion quantification using fuzzy-connectedness principles. IEEE Trans Med Imaging 1997;16:598 -609[Medline]
  26. Molyneux PD, Tubridy N, Parker GJ, et al. The effect of section thickness on MR lesion detection and quantification in multiple sclerosis. AJNR 1998;19:1715 -1720[Abstract]
  27. Gallagher HL, MacManus DG, Webb SL, Miller DH. A reproducible repositioning method for serial magnetic resonance imaging studies of the brain in treatment trials for multiple sclerosis. J Magn Reson Imaging 1997;7:439 -441[Medline]
  28. Hajnal JV, Saeed N, Soar EJ, Oatridge A, Young IR, Bydder GM. A registration and interpolation procedure for subvoxel matching of serially acquired MR images. J Comput Assist Tomogr 1995;19:289 -296[Medline]
  29. Tofts PS, Barker GJ, Filippi M, Gawne-Cain M, Lai M. An oblique cylinder contrast-adjusted (OCCA) phantom to measure the accuracy of MRI brain lesion volume estimation schemes in multiple sclerosis. Magn Reson Imaging 1997;15:183 -192[Medline]
  30. Herskovits EH, Itoh R, Melhem ER. Accuracy for detection of simulated lesions: comparison of fluid-attenuated inversion-recovery, proton density—weighted, and T2-weighted synthetic brain MR imaging. AJR 2001;176:1313 -1318[Abstract/Free Full Text]
  31. In den Kleef JJE, Cuppen JJM. RLSQ: T1, T2, and rho calculations, combining ratios and least squares. Magn Reson Med 1987;5:513 -524[Medline]
  32. Droege RT, Wiener SN, Rzeszotarski MS, et al. Nuclear magnetic resonance: a gray scale model for head images. Radiology 1983;148:763 -771[Abstract/Free Full Text]
  33. Sperber GO, Ericsson A, Hemmingsson A, Jung B, Thuomas KA. Improved formulae for signal amplitudes in repeated NMR sequences: applications in NMR imaging. Magn Reson Med 1986;3:685 -698[Medline]
  34. Melki PS, Jolesz FA, Mulkern RV. Partial RF echo planar imaging with the FAISE method. II. Contrast equivalence with spin-echo sequences. Magn Reson Med 1992;26:342 -354[Medline]
  35. Constable RT, Anderson AW, Zhong J, Gore JC. Factors influencing contrast in fast spin-echo MR imaging. Magn Reson Imaging 1992;10:497 -511[Medline]
  36. Melki PS, Mulkern RV, Panych LP, et al. Comparing the FAISE method with conventional dual-echo sequences. J Magn Reson Imaging 1991;1:319 -326[Medline]
  37. Constable RT, Gore JC. The loss of small objects in variable TE imaging: implications for FSE, RARE, and EPI. Magn Reson Med 1992;28:9 -24[Medline]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 RadiologyHome page
J. H. Woo, L. P. Henry, J. Krejza, and E. R. Melhem
Detection of Simulated Multiple Sclerosis Lesions on T2-weighted and FLAIR Images of the Brain: Observer Performance
Radiology, October 1, 2006; 241(1): 206 - 212.
[Abstract] [Full Text] [PDF]

Home page
 RadiologyHome page
L. Pikus, J. H. Woo, R. L. Wolf, E. H. Herskovits, G. Moonis, A. F. Jawad, J. Krejza, and E. R. Melhem
Artificial Multiple Sclerosis Lesions on Simulated FLAIR Brain MR Images: Echo Time and Observer Performance in Detection
Radiology, April 1, 2006; 239(1): 238 - 245.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Melhem, E. R.
Right arrow Articles by Itoh, R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Melhem, E. R.
Right arrow Articles by Itoh, R.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS