| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Pediatric Imaging |
1 From the Department of Pediatrics (B.J.D., T.L., V.J.S., L.K.), Department of Radiology (B.J.D., T.L., V.J.S., L.K., T.B.G.), Imaging Research Center (B.J.D., V.J.S.), and Division of Rheumatology (T.B.G.), Children’s Hospital Medical Center, 3333 Burnet Ave, Cincinnati, OH 45229. Received August 30, 2001; revision requested October 26; final revision received March 11, 2002; accepted March 25. Supported by an Arthritis Foundation clinical science grant. Address correspondence to B.J.D. (e-mail: bjd@athena.chmcc.org).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Twenty-five children and adolescents (age range, 5–17 years; mean age, 11.8 years) underwent a multisection-multiecho MR sequence for T2 relaxation time mapping. Quantitative transverse T2 maps of the patellar cartilage were calculated for 15 of the subjects. Sagittal T2 maps were calculated for the remaining 10 subjects. T2 profiles were generated for the patellar and distal femoral weight- and non–weight-bearing unossified epiphyseal and articular hyaline cartilage and for the distal femoral and proximal tibial physes. The Mann-Whitney U test was used to test for differences between paired profiles.
RESULTS: Femoral non–weight-bearing unossified epiphyseal and articular cartilage showed spatial variation similar to that of weight-bearing unossified epiphyseal and articular cartilage, but with increased T2 values (P < .001). T2 spatial variations of the distal femoral and proximal tibial physes were similar to those of epiphyseal and articular cartilage but had a different pattern and increased magnitude (P < .001). The highest T2 values were measured in the distal femoral physis.
CONCLUSION: T2 spatial variation of patellar hyaline cartilage in children is similar to that of patellar articular cartilage in adults. Mapping of spatial variation of T2 relaxation time of cartilage in the pediatric knee in vivo is feasible with a clinical 1.5-T MR imaging system and a body gradient coil.
© RSNA, 2002
Index terms: Arthritis, 45.771 • Arthritis, in infants and children, 45.771 • Children, skeletal system, 45.12146 • Knee, ligaments, menisci, and cartilage, 45.12146 • Knee, MR, 45.12141, 45.12146
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Prior work in adults has been performed with a 3.0-T MR imaging system (2,3,6,7) or a 1.5-T system with a specialized gradient coil (8). We hypothesized that unossified epiphyseal and articular cartilage in the pediatric knee demonstrates a pattern of T2 relaxation time variation similar to that of articular cartilage in the adult knee. Thus the purpose of our study was to determine the feasibility of mapping the spatial variation of cartilage T2 relaxation time in vivo in cartilaginous structures of the pediatric knee with a 1.5-T clinical MR imaging system and the manufacturer’s gradient coil.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Indications for MR imaging of the knee were the following: acute trauma (n = 10), swelling or known arthritis (n = 7), possible discoid meniscus (n = 4), pain (n = 2), possible metastatic disease (n = 1), and fibular osteochondroma (n = 1).
All MR imaging was performed with a 1.5-T unit (LX or Horizon; GE Medical Systems, Milwaukee, Wis) with a quadrature transmit-receive volume radio-frequency coil of 16 cm in diameter (IGC Medical Advances, Milwaukee, Wis). Institutional review board approval for the additional T2-relaxation-time imaging sequence was obtained. Patient informed consent was not required.
Imaging
A multisection-multiecho spin-echo MR imaging sequence (the parameters of which have been described previously [6]) was performed, from which T2 relaxation time profiles were calculated. Transverse images were obtained through the patella in 15 subjects (six of whom were male and nine female; age range, 8–17 years; mean age, 14.2 years). Sagittal images through the knee were obtained in 10 different subjects (five of whom were male and five female; age range, 5–15; mean age, 8.2 years). The plane of T2 relaxation time mapping was chosen randomly at the time of the examination. The younger, skeletally immature children were imaged in the sagittal plane to include the distal femoral and proximal tibial physes in the T2 relaxation time map.
Imaging parameters were as follows: repetition time msec/echo time msec, 1,500/9–99 in 9-msec increments for a total of 11 echo images; a 3-mm section thickness; a 1.5-mm gap; a 12-cm (transverse) or 16-cm (sagittal) field of view; a 256 x 160 matrix; and a 62.5-kHz receiver bandwidth. Total acquisition time was approximately 4 minutes. In the transverse plane, the phase-encoding axis was in the left-to-right direction to minimize the likelihood of pulsation artifact from the popliteal artery affecting the patellar cartilage. A 
-phase field of view in the anterior-to-posterior direction was used in the sagittal sequence to maximize resolution while decreasing imaging time to 3 minutes. Five sections were acquired in the transverse plane, and eight were acquired in the sagittal plane.
Data Analysis
T2 relaxation time maps.—T2 relaxation time maps were calculated on a pixel-by-pixel basis for all structures of the knee visualized in the transverse and sagittal planes by using a linear least squares curve-fitting algorithm at each section location. Image analysis was performed with Cincinnati Children’s Hospital Image Processing Software (CCHIPS), which is written in Interactive Data Language (IDL) (Research Systems, Boulder, Colo) (10,11). To minimize error of the regression, the initial spin-echo image was excluded. Echo images two through 11 with stimulated echo contribution were used for T2 relaxation time calculations. The influence of this error in the determination of in vivo T2 measurement has been discussed previously (6,12,13). Signal intensity as a function of time was fit to a monoexponential function for each pixel. A proton density map was generated from the pixel intercept data, and a T2 relaxation time map was generated from the slope of the best fit. The numeric values of the T2 relaxation time map were then represented in either a gray-scale or a color-coded image. Each pixel color represents an explicit range of T2 relaxation time values.
T2 relaxation time profiles.—T2 relaxation time profiles were calculated for the visualized cartilaginous structures in each plane. On transverse images, T2 relaxation time profiles were calculated for patellar cartilage. In younger children, patellar cartilage is a combination of articular and unossified cartilage, which are both forms of hyaline cartilage. On sagittal images, T2 relaxation time profiles were calculated for the cartilage of the patella, the distal femur (ie, weight-bearing and non–weight-bearing articular and unossified epiphyseal cartilage, all composed of hyaline cartilage), and the distal femoral and proximal tibial physes. Non–weight-bearing distal femoral cartilage was defined as that of the anterior third of the femoral condyle in the sagittal plane. Maximal weight-bearing distal femoral articular cartilage was defined as that of the middle third of the distal femoral condyle. Analysis of the posteriormost third was not performed because this cartilage may be both weight- and non–weight-bearing, depending on the degree of flexion of the knee.
Regions of interest (ROIs) were selected on T2 relaxation time images that displayed the largest amount of cartilage and were defined by a single operator (L.K.) (Fig 1). The CCHIPS/IDL computer program was used to automatically segment the cartilage from the surrounding bone on the basis of T2 relaxation time values and a K-means clustering algorithm (14,15). The operator defined the articulating surface of the cartilage. ROIs drawn around the distal femoral and proximal tibial physes were entirely computer-generated. Three points at a time were fit along the superiormost (or left) border of the computer-generated ROI. From this fit, a perpendicular line was then drawn to the inferiormost (or right) ROI boundary. The computer-generated default directions for the perpendicular lines ran from a superior-to-inferior and from a left-to-right direction for all ROIs. The T2 values calculated along this perpendicular line are referred to as the T2 relaxation time profiles.
|
Statistical evaluation.—ROIs were averaged over all subjects. Results are depicted as the mean T2 relaxation time (in milliseconds) ± standard error of the mean (SEM) across each group of subjects. The Mann-Whitney U test was used to determine statistical significance between average T2 relaxation time values for patellar cartilage imaged in the transverse plane and patellar cartilage imaged in the sagittal plane, for non-weight-bearing and weight-bearing distal femoral cartilage, and for the distal femoral and proximal tibial physes on a point-by-point basis. The level of significance was .05. For illustrative purposes, the averaged T2 relaxation time profile for each ROI was plotted on the same graph.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
|
|
|
For the normalized distance of 0.7–1.0, the average profiles in the sagittal plane were significantly different from the average profiles in the transverse plane (P < .05).
Femoral non–weight-bearing cartilage.—Sagittal views of the non–weight-bearing portion of femoral unossified epiphyseal and articular cartilage in 10 subjects generated 654 T2 relaxation time profiles, with a mean of 65.4 profiles per subject. The average cartilage thickness was 3.08 mm ± 1.14. The subchondral cartilage T2 relaxation time of 53 msec ± 1.9 decreased to a minimum of 52 msec ± 2.0 at 0.11 times the normalized distance and then monotonically increased to 85 msec ± 1.8 at the articular surface (normalized distance = 1) (Table, Fig 5).
|
Distal femoral physis.—The sagittal distal femoral physis measurements from 10 subjects resulted in 2,078 T2 relaxation profiles, with a mean of 207.8 profiles per subject. The average cartilage thickness was 1.61 mm ± 0.30. The T2 relaxation time at the metaphyseal border of 87 msec ± 1.5 decreased to a minimum of 70 msec ± 0.83 at 0.58 times the normalized distance and then monotonically increased to 77 msec ± 1.7 at the epiphyseal border (Table, Fig 6).
|
The average T2 relaxation time of the distal femoral physis was greater than the average T2 relaxation time profile through the proximal tibial physis (P < .001 for all normalized distances). The T2 relaxation time profiles for all interrogated regions are superimposed on the same graph (Fig 7). T2 relaxation time profiles calculated on the basis of identical images with swapped frequency-encoding directions indicated no effect from chemical shift artifact.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
T2 relaxation time mapping has been described with a 3-T commercial MR imager, as well as with a 1.5-T magnet that requires a specialized local gradient coil (2,3,6,8). Our study was successful with a 1.5-T clinical magnet with the manufacturer’s body gradient coil and commercially available extremity coil. Our patient population was younger than that observed in previous studies (2,3,6). The feasibility of using clinically available coils and the short imaging time of less than 4 minutes make this technique practical for imaging children.
T2 relaxation time maps of hyaline cartilage, which includes both unossified epiphyseal/apophyseal and articular cartilage, in children are similar to those obtained in adults for the patellar articular cartilage in the transverse plane (2,3,6). Children imaged in the sagittal plane were skeletally immature, and therefore we imaged both unossified epiphyseal (or patellar) cartilage and articular cartilage. We found that in both the transverse and sagittal planes there was a monotonic increase in T2 relaxation time from the subchondral bone toward the articular surface after a slight decrease at a normalized distance of 0.2–0.3 from the subchondral bone. The initial negative slope may be attributed to volume averaging with the subchondral bone (3). T2 relaxation time values of the subchondral bone are greater than those of cartilage, contributing to the artificial elevation of T2 values at a normalized distance of 0. The increase in the T2 relaxation time values at the subchondral bone boundary in the younger children may have resulted from volume averaging because the high-signal-intensity (ie, long T2 relaxation time) physis surrounds the patellar center of ossification (18).
The overall T2 relaxation time values obtained in the patellar cartilage in the transverse plane were lower than those obtained in the sagittal plane. We postulate that this discrepancy may relate to a 6-year mean age difference between the subjects imaged in the transverse plane and those imaged in the sagittal plane. However, to our knowledge, to date, no study has quantified age-related cartilage changes in children. The increased signal intensity of cartilage associated with more active ossification in younger children has been described by Varich et al (19). This is reflected as a quantitative increase in T2 relaxation time values in areas of more active cartilage growth and ossification. The range of T2 relaxation time values in both the transverse and sagittal planes is similar to those observed in the articular cartilage of the adult patella (2,3,6). Although all profiles were measured in an anterior-to-posterior direction within the patellar cartilage, the difference in T2 relaxation time values may be related to whether the data were acquired in a sagittal or transverse orientation. The angular dependence of T2 relaxation time in cartilage with the static magnetic field was not evaluated; however, it has been shown to contribute little to imaging of the adult distal femur and proximal tibia in vivo (7).
T2 relaxation time profiles of the weight-bearing and non–weight-bearing portions of the distal femur were evaluated separately. Unlike those in previous studies of adults, the profiles in our study incorporated both articular cartilage and unossified epiphyseal cartilage. All examinations were performed in children who had begun to walk. The biomechanical forces affecting the unossified femoral epiphyseal or patellar apophyseal cartilage are probably similar to those affecting the overlying articular cartilage. The weight-bearing cartilage demonstrated a T2 relaxation time profile similar to that of adult patellar articular cartilage. There was a monotonic increase toward the articular surface with an initial decrease that was probably attributable to volume averaging with the subchondral bone or with the physis of the distal femoral secondary center of ossification.
We chose the anterior aspect of the distal femoral condyle for our measurements of non–weight-bearing cartilage. This area has relatively less unossified epiphyseal cartilage and more homogeneous signal intensity than the posterior femoral condyle in children (19). The non–weight-bearing cartilage of the distal femur has a T2 relaxation time profile that increases from the subchondral bone to the articular surface. The overall T2 values are greater than those of the weight-bearing cartilage. This most likely corresponds to the known exudation of water from cartilage under compression (19–21). Because T2 relaxation time varies linearly with the water content, cartilage under biomechanical load will have a lower value (20,22,23).
A qualitative difference in the MR imaging appearance of the different regions of cartilage in the immature distal femoral condyle has been described by Varich et al (19) on sagittal fast- or conventional spin-echo T2-weighted and fast inversion-recovery–weighted MR images. Magic angle effects are thought to be minimal due to the less-organized distribution of collagen fibers in nonarticular hyaline cartilage (19). The patients in their study group were younger (median age, 27 months) than the children we evaluated. However, our quantitative results are in agreement with their observations.
T2 relaxation time profiles were successfully obtained from the distal femoral and proximal tibial physis in children. The profiles showed more of a "U" shape than those obtained in epiphyseal cartilage (Fig 7). The nadir of T2 relaxation time values was seen approximately midway between the metaphyseal and epiphyseal borders. This may be partially due to volume averaging on either side of the physis (ie, from the metaphyseal bone and from the secondary center of ossification). The thickness of the interrogated physeal cartilage was approximately 1.5 mm. The limited pixel count used in the determination of the T2 profile may affect the shape of the spatial variation. The minimal T2 values obtained in the physeal cartilage are greater than those obtained in the epiphyseal cartilage (Fig 7). This presumably reflects the difference in water mobility. Water is more tightly bound in epiphyseal cartilage than in physeal cartilage (24,25). This results in a relative overall decrease in T2 relaxation time of the epiphyseal cartilage.
The distal femur accounts for 57% and the tibia accounts for 43% of longitudinal growth at the knee (26,27). The distal femoral physis had greater T2 relaxation time values than the proximal tibial physis over the whole profile (P < .001). This relatively greater rate of endochondral ossification at the physis of the distal femur over the proximal tibia may be reflected in greater T2 values (19,28). Thus, the dynamics of normal skeletal growth are illustrated in increased T2 relaxation time.
The limitations of our study included a lack of histologic correlation with T2 relaxation time profiles in children. Our patient population was imaged for determining the feasibility of the technique, and patients were included based on our ability to perform an additional imaging sequence without the need for additional sedation. We did not image control subjects, nor did we base inclusion in the study on clinical diagnosis. Potential alterations in T2 relaxation time maps might result from underlying pathologic conditions. However, a patient was not included if the pathologic condition was seen to affect the epiphyseal or physeal cartilage of the knee with conventional imaging sequences. Some patients with known arthritis were imaged. T2 relaxation time maps were not evaluated for focal cartilaginous defects. The study group was too small to correct for sex or patient age. All patients imaged in the sagittal plane were skeletally immature so that the physes around the knee could be evaluated. We did not exclude skeletally mature patients from our transverse imaging protocol. Future studies in which child volunteers with varying degrees of skeletal maturity are imaged are necessary so that conclusions relating to normal growth can be drawn. We performed calculations on sections that included the thickest cartilage to diminish volume averaging with bone. Therefore, thinner regions of cartilage were not sampled. We combined the T2 relaxation time measurements of unossified epiphyseal and apophyseal cartilage with those of articular cartilage because the articular cartilage in small children is very thin. Because our study was directed at feasibility, we did not differentiate between these two types of hyaline cartilage.
In this study, we performed T2 relaxation time mapping in the pediatric knee with a clinical 1.5-T MR system to generalize patterns of T2 spatial variation in the knees of children. This technique does not require the use of intravenous or intraarticular contrast medium, and no repeated or delayed image acquisition is necessary. This technique, however, does not result in direct imaging of proteoglycan concentration but probably reflects the interaction of water with the cartilage matrix. Further work is required to determine the spatial T2 relaxation time variation in healthy control children who have no clinical symptoms around the knee and to determine the contributions to spatial T2 variation of age, skeletal maturation, and sex.
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, B.J.D., T.L.; study concepts, B.J.D., T.L., V.J.S., T.B.G.; study design, B.J.D., T.L.; literature research, B.J.D., T.L.; clinical studies, B.J.D., T.L., T.B.G.; experimental studies, B.J.D., T.L.; data acquisition, B.J.D., T.L., L.K.; data analysis/interpretation, all authors; statistical analysis, B.J.D., V.J.S., T.B.G.; manuscript preparation, B.J.D., T.L.; manuscript definition of intellectual content, all authors; manuscript editing, B.J.D., T.L., T.B.G.; manuscript revision/review and final version approval, all authors.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. Mascalchi, G. Belli, M. Zappa, G. Picozzi, M. Falchini, R. D. Nave, G. Allescia, A. Masi, A. L. Pegna, N. Villari, et al. Risk-benefit analysis of X-ray exposure associated with lung cancer screening in the Italung-CT trial. Am. J. Roentgenol., August 1, 2006; 187(2): 421 - 429. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. C. Essman, M. R. Lewis, and W. H. Miller Intraorgan Biodistribution and Dosimetry of 153Sm-Ethylenediaminetetramethylene Phosphonate in Juvenile Rabbit Tibia: Implications for Targeted Radiotherapy of Osteosarcoma J. Nucl. Med., December 1, 2005; 46(12): 2076 - 2082. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Watrin-Pinzano, J.-P. Ruaud, P. Olivier, L. Grossin, P. Gonord, A. Blum, P. Netter, G. Guillot, P. Gillet, and D. Loeuille Effect of Proteoglycan Depletion on T2 Mapping in Rat Patellar Cartilage Radiology, January 1, 2005; 234(1): 162 - 170. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Van Breuseghem, H. T. C. Bosmans, L. V. Elst, F. Maes, S. D. Pans, P. P. M. A. Brys, E. A. Geusens, and G. J. Marchal T2 Mapping of Human Femorotibial Cartilage with Turbo Mixed MR Imaging at 1.5 T: Feasibility Radiology, November 1, 2004; 233(2): 609 - 614. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. A. Connolly, D. Jaramillo, J. K. Hong, and F. Shapiro Skeletal Development in Fetal Pig Specimens: MR Imaging of Femur with Histologic Comparison Radiology, November 1, 2004; 233(2): 505 - 514. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Lazovic-Stojkovic, T. J. Mosher, H. E. Smith, Q. X. Yang, B. J. Dardzinski, and M. B. Smith Interphalangeal Joint Cartilage: High-Spatial-Resolution in Vivo MR T2 Mapping--A Feasibility Study Radiology, October 1, 2004; 233(1): 292 - 296. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. McWilliams, J. Mayo, S. MacDonald, J. C. leRiche, B. Palcic, E. Szabo, and S. Lam Lung Cancer Screening: A Different Paradigm Am. J. Respir. Crit. Care Med., November 15, 2003; 168(10): 1167 - 1173. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
