| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Obstetric Imaging |
1 From the Department of Radiology (G.W., F.V.C., A.Q., B.N.J., R.A.F.) and the Fetal Treatment Center, Department of Surgery (D.L.F.), University of California San Francisco, 505 Parnassus Ave, Box 0628, San Francisco, CA 94143-0628. From the 2003 RSNA scientific assembly. Received September 26, 2003; revision requested December 5; revision received February 12, 2004; accepted March 23. Address correspondence to F.V.C. (e-mail: fergus.coakley@radiology.ucsf.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Total lung volume (TLV) was measured by using planimetry on single-shot rapid acquisition with relaxation enhancement magnetic resonance (MR) images obtained in 91 fetuses with ultrasonographically (US) normal chests and 28 fetuses with US-determined pulmonary hypoplasia. All fetuses were aged between 18 and 38 weeks gestation. Analysis of covariance was used to identify parameters that were not different between the fetuses with US-determined normal and those with US-determined abnormal chests, and these variables were used to construct an algorithm for calculating predicted lung volume. The relative lung volume—that is, the observed lung volume expressed as a percentage of the predicted lung volume—was then calculated in fetuses with pulmonary hypoplasia.
RESULTS: There was no significant difference in mean maternal or gestational age between the two fetus groups. Stepwise regression analysis was used to generate the following equation for predicting fetal lung volume on the basis of independent biometric indexes, with a correlation coefficient of 0.93: TLV = (0.52 · LV) + (0.33 · BD) – (0.06 · FL) – 13.7, with TLV and liver volume (LV) in milliliters and biparietal diameter (BD) and femoral length (FL) in centimeters. In the fetuses with normal chests, relative lung volume varied between 51% and 134%. In the fetuses with pulmonary hypoplasia, relative lung volume varied between 6% and 70%.
CONCLUSION: The predicted lung volume in fetuses of a wide range of gestational ages can be calculated with a high degree of accuracy, enabling prenatal MR imaging lung volumetry in which relative lung volume is used to quantify fetal pulmonary hypoplasia.
© RSNA, 2004
Index terms: Fetus, abnormalities, 856.8754, 856.8755 • Fetus, MR, 856.121412, 856.121416 • Fetus, respiratory system • Lung, ventilation
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
In our initial study (1), we evaluated data on 24 fetuses with normal chests and gestational ages of 18–29 weeks. The small number of fetuses and the limited range of gestational ages were considered potential limitations to the wider use of the proposed algorithm, particularly later in gestation when lung volumes appear to be more variable (4). In addition, in that study we did not investigate whether the biometric measurements used to determine predicted lung volume were independent of the grouping of fetuses as study subjects or control subjects. Therefore, we performed the current study to retrospectively determine a biometric algorithm for calculating relative lung volume in fetuses with normal lungs and a wide range of gestational ages by using proved independent variables and to retrospectively investigate the use of this algorithm in fetuses with pulmonary hypoplasia.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
The following exclusion criteria were applied: (a) Visualization of the lungs at MR imaging was incomplete in 18 cases. Many of the examinations were performed for neurologic indications, and, thus, the lungs were not necessarily included in the field of view. (b) US depicted an extrathoracic abnormality, such as gross hydrocephalus, that might have confounded the biometric measurements of fetal size in five cases. The final control group consisted of 91 fetuses in 85 pregnancies: 77 singleton pregnancies, six twin pregnancies with both fetuses included, and two twin pregnancies with one fetus each included. The mean maternal age for the control group was 30 years ± 5.9 (standard deviation) (maternal age range, 17–47 years). The mean gestational age was 24 weeks ± 4 (gestational age range, 18–38 weeks).
Indications for MR imaging were spinal anomalies such as myelomeningocele (n = 25), isolated mild lateral cerebral ventriculomegaly (n = 18), complicated twin pregnancy (n = 13), suspected extrathoracic structural anomaly (n = 10), low-grade intracranial hemorrhage (n = 7), other suspected neurologic abnormality (n = 5), suspected placenta accreta (n = 3), abdominal mass (n = 2), neck mass (n = 2), amniotic band syndrome (n = 1), suspected intrauterine growth restriction (n = 1), sacrococcygeal teratoma (n = 1), cardiac calcification (n = 1), suspected tuberous sclerosis (n = 1), and a small subcutaneous occipital fluid collection (n = 1). None of the neurologic abnormalities was associated with cranial enlargement.
We identified the group of fetuses with pulmonary hypoplasia by reviewing the records of all patients referred for obstetric MR imaging between 1996 and 2002 and including those cases in which the results of contemporaneous obstetric US that had been performed at our institution suggested abnormally small lung volumes related to unilateral congenital diaphragmatic hernia (n = 24), bilateral congenital diaphragmatic hernia (n = 1), thoraco-omphalopagus twins (n = 2), or severe oligohydramnios (n = 1). The group of fetuses with pulmonary hypoplasia consisted of 28 fetuses in 27 pregnancies: 26 singleton pregnancies and one twin pregnancy. The mean maternal age in this group was 30 years ± 6.1 (material age range, 17–47 years). The mean gestational age was 25 weeks ± 3 (gestational age range, 21–36 weeks). Twenty-four fetuses in the control group and 22 fetuses in the pulmonary hypoplasia group were included in our previous study of fetal lung volumetry (1).
US Technique
US was performed by using state-of-the-art equipment (Sequoia; Acuson, Mountain View, Calif) with 3.5–8.0-MHz transducers and combined gray-scale and color Doppler examinations. All US images were reviewed by one of four attending radiologists (including R.A.F.), each of whom had at least 10 years experience in prenatal diagnosis, who also reported the findings. US biometry was performed by using standard described methodology (5) and included measurements of biparietal diameter, femoral length, abdominal circumference, head circumference, and estimated fetal weight.
MR Imaging Technique
MR imaging was performed on the same day as US in 67 fetuses, between 1 and 4 days after US in 22 fetuses, and between 5 and 22 days after US in 30 fetuses. MR imaging was performed with a 1.5-T superconducting whole-body unit (Signa; GE Medical Systems, Milwaukee, Wis) by using a four-element phased-array surface coil (GE Medical Systems). T1-weighted MR images were obtained by using a breath-hold spoiled gradient-echo sequence and the following parameters: 100–150/4.2 (repetition time msec/echo time msec), a flip angle of 70°–90°, a section thickness of 5–7 mm, an intersection gap of 0–1 mm, a matrix of 256 x 128–192, and one acquired signal.
T2-weighted MR images were obtained by using a single-shot rapid acquisition with relaxation enhancement sequence with 
/90, a section thickness of 4–6 mm, an intersection gap of 0 mm, a matrix of 256 x 160–256, and one partial Fourier signal acquisition. The supervising radiologist optimized the field of view, the number of sections acquired, the section thickness, and the intersection gap for each fetus. Sequences that were degraded by fetal motion were repeated, with the technical parameters altered, if necessary, to produce shorter acquisition times. A variable bandwidth was used with all sequences. The acquisition time was 15–30 seconds.
MR Imaging Volumetry
A postdoctoral scholar (G.W.) was trained by the principal investigator in our earlier study (F.V.C.) to perform planimetric measurements of the fetal lung and the fetal liver. Lung volume was measured by selecting the single-shot rapid acquisition with relaxation enhancement sequence that enabled imaging of both lungs in their entirety without motion or misregistration artifacts (Fig 1a). On each section, the cross-sectional area of the lung was measured by using the free-form region-of-interest tool on a picture archiving and communication system (Impax; Agfa-Gevaert, Mortsel, Belgium). The area was multiplied by the section thickness to determine the volume of that section. The volumes of all sections were summed to determine the volume of the entire lung. This calculation was repeated for the contralateral lung, and the volumes of the right and left lungs were combined to determine the total lung volume.
|
|
Liver volume could not be measured in nine fetuses: six fetuses without and three fetuses with pulmonary hypoplasia. The time required to perform these measurements ranged from 15 to 45 minutes per case, depending on the fetal size, the fetal orientation, and the number of T2-weighted sequences used to average the measurements.
Statistical Analyses
With use of established growth curves, the baseline US biometric measurements obtained in the 30 fetuses in whom the interval between US and MR imaging was longer than 4 days were adjusted to reflect the developmental change during that interval (5). That is, the measurements were adjusted to the value that corresponded to the same percentile value at the date of MR imaging. Student t tests were performed to compare mean maternal and gestational ages between the fetuses with normal chests and those with pulmonary hypoplasia and to compare the mean intervals between US and MR imaging between the two groups. Analysis of covariance was used to compare US (head circumference, biparietal diameter, femoral length, and estimated fetal weight) and MR imaging (liver volume) biometric parameter measurements between the fetuses with normal chests and those with pulmonary hypoplasia.
Multivariate regression analysis of associations between total lung volume, gestational age, and the imaging parameters found to be independent in the analysis of covariance was then performed to generate algorithms for predicting lung volume. These algorithms were tested for correlations with observed total lung volumes in the fetuses with normal lungs and were used to calculate the relative lung volume—that is, the observed lung volume expressed as a percentage of the predicted lung volume—in the fetuses with normal chests and in those with pulmonary hypoplasia.
Of the total of 119 data sets, 110 were complete: 85 of 91 data sets for the control group and 25 of 28 data sets for the pulmonary hypoplasia group. Nine data sets had missing liver volume data: six of the 91 control group data sets and three of the 28 pulmonary hypoplasia group data sets. Statistical calculations were performed by using a statistical software package (Statistica 5.1; StatSoft, Tulsa, Okla) and computer software (Excel 2002; Microsoft, Redmond, Wash). P 
.05 was considered to indicate statistical significance.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
|
When only US indexes were used, the following equation to predict fetal lung volume was generated at stepwise regression analysis: TLV = (0.30 · HC) – (0.12 · BD) + (0.40 · FL) – 46.6. The correlation coefficient for the relationship between total lung volume, as predicted by using this equation, and observed total lung volume in the fetuses with normal chests was 0.85.
The scatterplot in Figure 2 shows the relationship between observed and predicted total lung volumes in the two fetus groups, including cases in which liver volumetry was (in 85 of the 91 fetuses without and 25 of the 28 fetuses with pulmonary hypoplasia) and was not (in six of the 85 fetuses without and three of the 28 fetuses with pulmonary hypoplasia) possible. These data are presented in another scatterplot in Figure 3.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Second, we included fetuses of a wide range of gestational ages, from 18 to 38 weeks, which is probably the gestational period during which measurements of relative lung volume might be clinically useful. We were motivated to validate our algorithm in fetuses of a wide range of gestational ages because lung volume appears to become more variable as gestation progresses (4).
Third, we used analysis of covariance to compare the biometric parameters used in the algorithm for predicting lung volume between fetuses with normal chests and those with pulmonary hypoplasia to establish that these parameters were independent of the assignments of fetuses to control versus pulmonary hypoplasia groups. We found lung volume to be the only parameter that was significantly different between the two groups. This is a crucial requirement for the use of these parameters in the algorithm for predicting relative lung volume because if these parameters were systematically different in the pulmonary hypoplasia group, the calculation of predicted lung volume would be unreliable. For example, reduced abdominal circumference is said to be associated with herniation of the abdominal viscera into the chest, so the incorporation of this variable into an algorithm for calculating predicted lung volume might be misleading (6–8).
Our study complements previous research of the volumetric assessment of fetal lung growth. We used stepwise linear regression analysis to generate an algorithm for predicting fetal lung volume with a correlation coefficient of 0.93. Linear regression was considered an appropriate analysis method because increases in similar biometric indicators—specifically, abdominal circumference, biparietal diameter, femoral length, and head circumference—are known to be linear (9–11) and because increased fetal lung volume has also been shown to be linear or near linear in the 18–38-week gestational age range (12).
Other authors have investigated the use of three-dimensional US for volumetry of fetal organs. Three-dimensional US estimations of fetal lung volumes (13) are consistently lower than volumes measured at MR imaging planimetry. This may reflect the poor acoustic contrast between the lung, the liver, and adjacent structures at US (14–16). In contrast, the fetal lungs and the fetal liver are well demarcated on T2- and T1-weighted MR images, respectively, facilitating accurate planimetry (17–19). In a recent study to investigate tracheal occlusion in a group of seven sheep fetuses, a correlation coefficient of 0.84 for the relationship between the lung volume measured at true fast steady-state free precession MR imaging and the lung volume measured at autopsy was derived (20).
Our study had a number of limitations. First, our control group of fetuses with normal chests was from a population referred for prenatal MR imaging because of various extrathoracic abnormalities and therefore may not have been representative of truly healthy fetuses. We tried to minimize the effect of this selection process by excluding fetuses with thoracic abnormalities at US and those with US abnormalities that were likely to confound biometric measurements.
Second, the pulmonary hypoplasia group consisted predominantly of fetuses with congenital diaphragmatic hernia. The inclusion of a wider spectrum of abnormalities associated with pulmonary hypoplasia would have strengthened the part of our study in which we investigated the use of relative lung volume measurements in fetuses with pulmonary hypoplasia.
Third, a single observer performed all of the planimetric measurements. However, these measurements are relatively straightforward, as confirmed in our previous work (2), in which there was excellent interobserver agreement on measurements of fetal lung volume (intraclass correlation coefficient, 0.95).
Fourth, our contention that relative lung volume measurements are useful for evaluating pulmonary hypoplasia raises the question of which relative lung volume threshold should be used to define pulmonary hypoplasia. In our study, however, the limited range of abnormalities in the pulmonary hypoplasia group made answering this question problematic. One approach, based on our finding of a 17% standard deviation in relative lung volume in the fetuses with normal chests, would be to consider fetuses with a relative lung volume of less than 64% (more than 2 standard deviations below the normal value) to have pulmonary hypoplasia. Although the use of this approach seems reasonable and would have enabled the correct identification of all but one of the fetuses with pulmonary hypoplasia, with only three false-positive results, we are reluctant to recommend the use of such a diagnostic criterion until we have examined more fetuses with pulmonary hypoplasia related to causes other than congenital diaphragmatic hernia.
Fifth, in this study we did not correlate relative lung volume with postnatal outcome. Although our previous study results demonstrated that a relative lung volume of less than 40% is associated with a poorer prognosis for fetuses with congenital diaphragmatic hernia (2), more outcome studies of other causes of pulmonary hypoplasia are required. It is plausible that relative lung volume is only one factor that affects the outcome of such fetuses; underlying abnormality and functional lung maturity also are factors that probably influence outcome. For example, clinicians involved in the care of infants with congenital anomalies are aware that pulmonary compression due to congenital cystic adenomatoid malformation appears to be associated with less morbidity than an equivalent degree of pulmonary compression due to congenital diaphragmatic hernia.
Sixth, because of difficulties in monitoring the fetal cardiac cycle and chest wall motion, we did not investigate whether or how the cardiac cycle or fetal breathing movements might affect fetal lung volume measurements. Presumably, such effects would have been averaged across the acquisition time and by performing multiple measurements during numerous acquisitions to generate a mean fetal lung volume value.
In conclusion, predicted lung volume can be calculated with a high degree of accuracy in fetuses of a wide range of gestational ages to enable prenatal MR imaging lung volumetry in which the determined relative lung volume is used to quantify fetal pulmonary hypoplasia.
|
FOOTNOTES |
|---|
Authors stated no financial relationship to disclose.
Author contributions: Guarantor of integrity of entire study, G.W.; study concepts, F.V.C.; study design, G.W., F.V.C.; literature research, G.W., F.V.C.; clinical studies, F.V.C., A.Q.; data acquisition, all authors; data analysis/interpretation, G.W.; statistical analysis, all authors; manuscript preparation, definition of intellectual content, editing, revision/review, and final version approval, all authors
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. A. Busing, A. K. Kilian, T. Schaible, D. J. Dinter, and K. W. Neff MR Lung Volume in Fetal Congenital Diaphragmatic Hernia: Logistic Regression Analysis--Mortality and Extracorporeal Membrane Oxygenation Radiology, July 1, 2008; 248(1): 233 - 239. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. A. Busing, A. K. Kilian, T. Schaible, C. Endler, R. Schaffelder, and K. W. Neff MR Relative Fetal Lung Volume in Congenital Diaphragmatic Hernia: Survival and Need for Extracorporeal Membrane Oxygenation Radiology, July 1, 2008; 248(1): 240 - 246. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. M. Cannie, J. C. Jani, F. Van Kerkhove, J. Meerschaert, F. De Keyzer, L. Lewi, J. A. Deprest, and S. Dymarkowski Fetal Body Volume at MR Imaging to Quantify Total Fetal Lung Volume: Normal Ranges Radiology, April 1, 2008; 247(1): 197 - 203. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. A. Busing, A. K. Kilian, T. Schaible, A. Debus, C. Weiss, and K. W. Neff Reliability and Validity of MR Image Lung Volume Measurement in Fetuses with Congenital Diaphragmatic Hernia and in Vitro Lung Models Radiology, December 1, 2007; 246(2): 553 - 561. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Gorincour, D. Eurin, F. E. Avni, For the GRRIF Group (Groupe Radiopediatrique de Re, M. Cannie, J. C. Jani, F. De Keyzer, S. Dymarkowski, and J. A. Deprest Prenatal Prediction of Pulmonary Hypoplasia: US and MR Imaging Working Together Radiology, November 1, 2007; 245(2): 608 - 609. [Full Text] [PDF]
|
|
|
|
|
|
J. C. Jani, M. Cannie, C. F. A. Peralta, J. A. Deprest, K. H. Nicolaides, and S. Dymarkowski Lung Volumes in Fetuses with Congenital Diaphragmatic Hernia: Comparison of 3D US and MR Imaging Assessments Radiology, August 1, 2007; 244(2): 575 - 582. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. F. Kazan-Tannus, V. Dialani, M. L. Kataoka, G. Chiang, H. A. Feldman, J. S. Brown, and D. Levine MR Volumetry of Brain and CSF in Fetuses Referred for Ventriculomegaly Am. J. Roentgenol., July 1, 2007; 189(1): 145 - 151. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. L. Ward, M. Nishino, H. Hatabu, J. A. Estroff, C. E. Barnewolt, H. A. Feldman, and D. Levine Fetal Lung Volume Measurements: Determination with MR Imaging--Effect of Various Factors. Radiology, July 1, 2006; 240(1): 187 - 193. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. F. A. Peralta, J. F. Kazan-Tannus, V. Bunduki, E. M. Santos, C. C. de Castro, G. G. Cerri, and M. Zugaib Evaluation of the Agreement Between 3-Dimensional Ultrasonography and Magnetic Resonance Imaging for Fetal Lung Volume Measurement J. Ultrasound Med., April 1, 2006; 25(4): 461 - 467. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. F. Goncalves, W. Lee, J. Espinoza, and R. Romero Three- and 4-Dimensional Ultrasound in Obstetric Practice: Does It Help? J. Ultrasound Med., December 1, 2005; 24(12): 1599 - 1624. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
) and 28 fetuses with pulmonary hypoplasia (
). Predicted lung volumes were calculated by using an algorithm based on liver volume, biparietal diameter, and femoral length measurements in 85 fetuses without and 25 fetuses with pulmonary hypoplasia. For the nine remaining fetuses (six of 91 fetuses without and three of 28 fetuses with pulmonary hypoplasia), liver volume measurements were not available, so an algorithm based on head circumference, biparietal diameter, and femoral length measurements was used to calculate the predicted lung volume.