MRI investigation of normal fetal lung maturation using signal intensities on different imaging sequences

Fetal MRI radiomics: non-invasive and reproducible quantification of human lung maturity

OBJECTIVES

To assess the reproducibility of radiomics features extracted from the developing lung in repeated in-vivo fetal MRI acquisitions.

METHODS

In-vivo MRI (1.5 Tesla) scans of 30 fetuses, each including two axial and one coronal T2-weighted sequences of the whole lung with all other acquisition parameters kept constant, were retrospectively identified. Manual segmentation of the lungs was performed using ITK-Snap. One hundred radiomics features were extracted from fetal lung MRI data using Pyradiomics, resulting in 90 datasets. Intra-class correlation coefficients (ICC) of radiomics features were calculated between baseline and repeat axial acquisitions and between baseline axial and coronal acquisitions.

RESULTS

MRI data of 30 fetuses (12 [40%] females, 18 [60%] males) at a median gestational age of 24 + 5 gestational weeks plus days (GW) (interquartile range [IQR] 3 + 3 GW, range 21 + 1 to 32 + 6 GW) were included. Median ICC of radiomics features between baseline and repeat axial MR acquisitions was 0.92 (IQR 0.13, range 0.33 to 1), with 60 features exhibiting excellent (ICC > 0.9), 27 good (> 0.75-0.9), twelve moderate (0.5-0.75), and one poor (ICC < 0.5) reproducibility. Median ICC of radiomics features between baseline axial and coronal MR acquisitions was 0.79 (IQR 0.15, range 0.2 to 1), with 20 features exhibiting excellent, 47 good, 29 moderate, and four poor reproducibility.

CONCLUSION

Standardized in-vivo fetal MRI allows reproducible extraction of lung radiomics features. In the future, radiomics analysis may improve diagnostic and prognostic yield of fetal MRI in normal and pathologic lung development.

KEY POINTS

• Non-invasive fetal MRI acquired using a standardized protocol allows reproducible extraction of radiomics features from the developing lung for objective tissue characterization. • Alteration of imaging plane between fetal MRI acquisitions has a negative impact on lung radiomics feature reproducibility. • Fetal MRI radiomics features reflecting the microstructure and shape of the fetal lung could complement observed-to-expected lung volume in the prediction of postnatal outcome and optimal treatment of fetuses with abnormal lung development in the future.

Comparison of fetal lung maturation in fetuses with intrauterine growth restriction with control group, using lung volume, lung/liver and lung/muscle signal intensity and apparent diffusion coefficient ratios on different magnetic resonance imaging sequences

PURPOSE

To compare lung volume, lung apparent diffusion coefficient (ADC) and signal intensity ratio (SIR) on different magnetic resonance imaging (MRI) sequences between intrauterine growth restriction (IUGR) fetuses and the control group.

MATERIALS AND METHODS

49 IUGR and 58 non-IUGR fetuses were imaged using 3 Tesla MRI units. Total lung volume (TLV), lung/liver SIR (LLSIR) and lung/muscle SIR (LMSIR) in T1 and T2-weighted sequences and lung/liver ADC ratio (LLADCR) and lung/muscle ADC ratio (LMADCR) were assessed.

RESULTS

LLSIR and LMSIR were significantly higher in the T1-weighted sequence (p-value: .03) and LLADCR and LMADCR were significantly lower on diffusion-weighted imaging (DWI) in IUGR fetuses compared to the control group (p-value: .01). There was no significant difference in SIRs in the T2-weighted sequence between the two groups. Although TLV was increased with gestational age in both groups, it was significantly lower in the IUGR group (mean: 82 ± 22.7 ml vs. 110.8 ± 18 ml, p-value: <.001).

CONCLUSION

The T1-weighted sequence and DWI seem to be better than the T2-weighted sequence for assessing the faint difference of lung maturity between groups. However, SIR differences were not as meaningful as TLV differences and this could be related to the complex maturation process in IUGR fetuses as the effect of higher endogenous corticosteroids.

Prenatal evaluation of functional pulmonary hypoplasia via fetal magnetic resonance imaging

OBJECTIVE

The purpose of this study was to retrospectively examine the use of lung-to-liver signal intensity ratio (LLSIR) on T2-weighted images to predict functional pulmonary hypoplasia.

METHODS

The subjects of this study were pregnant women who underwent magnetic resonance imaging (MRI). Patients who required nitric oxide inhalation and those who died from respiratory disorders were classified as having functional pulmonary hypoplasia (FPH). All other cases were presented as the control group. We retrospectively analyzed MRI and perinatal data. LLSIR was defined as the ratio of lung signal intensity to liver signal intensity. We examined the relationship between LLSIR and gestational age, compared the LLSIRs in the two groups, and calculated the best cut-off value of the LLSIR to predict FPH.

RESULTS

One hundred and ninety-one patients were eligible for this study, and 12 cases were classified as having FPH. In the control group, LLSIR increased with age (r = 0.383, p < 0.001). We used the observed/expected LLSIR (o/e LLSIR), which was the ratio of obtained LLSIR to expected LLSIR calculated by the regression line to correct the effect of gestational age. In the FHP group, o/e LLSIR was significantly lower than in the control group (p < 0.001). A receiver operating characteristic curve analysis showed that cases with o/e LLSIR above 0.85 were less likely to cause FPH.

CONCLUSIONS

Low o/e LLSIR might reflect the histological characteristics of hypoplastic lung structures. O/e LLSIR seems to be a useful MRI parameter for screening FPH.

IntraVoxel Incoherent Motion (IVIM) MRI of fetal lung and kidney: Can the perfusion fraction be a marker of normal pulmonary and renal maturation?

PURPOSE

To investigate the use of IntraVoxel Incoherent Motion (IVIM) MRI in the study of microstructural tissue changes occurring in fetal lung and kidney during gestation.

METHODS

34 normal pregnancies were enrolled. Patients were divided into two groups based on gestational age (GA): group A (21-29 weeks) and group B (30-39 weeks). MR examinations were performed at 1.5T, with a standard fetal MR protocol including a Diffusion-Weighted Echo-Planar Imaging sequence with 10 different b-values (0, 10, 30, 50, 75, 100, 200, 400, 700, 1000s/mm²). For each fetus, two bilateral ROIs were manually placed in lung and renal parenchyma. Mean values of perfusion fraction f, pseudo-diffusion coefficient D* and diffusion coefficient D were obtained. The correlation between IVIM parameters and GA was investigated.

RESULTS

In renal ROIs a positive correlation between f_kidney and GA (p < 0.005) was found; similarly f_lung showed a statistically significant correlation with GA (p < 0.001). F mean values were significantly higher in group B compared to group A in both renal (p = 0.0002) and lung (p = 0.018) ROIs. No correlation was found in D and D* as a function of GA.

CONCLUSIONS

The IVIM perfusion fraction f may be considered as a potential marker of pulmonary and renal maturation in relation to hemodynamic changes described in intrauterine life. Our results highlight that IVIM model is useful as an additional prenatal diagnostic tool to study lung and renal development.

What is the most suitable MR signal index for quantitative evaluation of placental function using Half-Fourier acquisition single-shot turbo spin-echo compared with T2-relaxation time?

Background Half-Fourier acquisition single-shot turbo spin-echo (HASTE) imaging is now widely used for placental and fetal imaging because of its rapidity and low sensitivity to fetal movement. If placental dysfunction is also predicted by quantitative value obtained from HASTE image, then it might be beneficial for evaluating placental wellbeing. Purpose To ascertain the most suitable magnetic resonance (MR) signal indexes reflecting placental function using HASTE imaging. Material and Methods This retrospective study included 37 consequent patients who had given informed consent to MR imaging (MRI) examinations. All had undergone MRI examinations between February 2014 and June 2015. First, the correlation between T2-relaxation time of normal placenta and gestational age (GA) was examined. Second, correlation between signal intensity ratios (SIRs) using HASTE imaging and placental T2-relaxation time were assessed. The SIRs were calculated using placental signal intensity (SI) relative to the SI of the amniotic fluid, fetal ocular globes, gastric fluid, bladder, maternal psoas major muscles, and abdominal subcutaneous adipose tissue. Results Among the 37 patients, the correlation between T2-relaxation time of the 25 normal placentas and GA showed a moderately strong correlation (Spearman rho = -0.447, P = 0.0250). The most significant correlation with placental T2-relaxation time was observed with the placental SIR relative to the maternal psoas major muscles (SIR_{pl./psoas muscle}) (Spearman rho = -0.531, P = 0.0007). Conclusion This study revealed that SIR_{pl./psoas muscle} showed the best correlation to placental T2-relaxation time. Results show that SIR_{pl./psoas muscle} might be optimal as a clinically available quantitative index of placental function.

The fetal lung-to-liver signal intensity ratio on magnetic resonance imaging as a predictor of outcomes from isolated congenital diaphragmatic hernia

PURPOSE

We investigated the developmental changes in the unaffected contralateral lungs of patients with isolated left-sided congenital diaphragmatic hernia (CDH) using signal intensity ratios on prenatal magnetic resonance imaging (MRI) and determined whether these changes correlated with clinical outcomes.

METHODS

We performed 47 fetal MRI screens on 30 patients with isolated left-sided CDH. A cohort of 88 fetuses was selected as the control. We calculated the lung-to-liver signal intensity ratio (LLSIR) using region of interest analysis and compared LLSIR between the groups and between those in the CDH group with good and poor prognoses.

RESULTS

In the control group, LLSIR increased as pregnancy progressed [regression line = 2.232 + 0.135 × (GW-23), r = 0.669]. In the CDH group, especially in the poor prognosis group, LLSIR did not significantly increase as pregnancy progressed [regression line for good prognosis = 1.827 + 0.092 × (gestational week-23), r = 0.733; regression line for poor prognosis = 1.731 + 0.025 × (gestational week-23), r = 0.634].

CONCLUSION

Fetal LLSIR on T2-weighted MRI is an accurate marker of fetal lung maturity that correlates with postnatal survival and can potentially be used as a prognostic parameter in CDH management.

Imaging findings of bronchial atresia in fetuses, neonates and infants

Congenital lung malformations are increasingly detected before birth. However, bronchial atresia is rarely identified in utero and not always recognized in neonates. There are two types of atresia: 1) proximal, located at the level of the mainstem or the proximal lobar bronchi, which is extremely rare and usually lethal during pregnancy, causing a tremendous volume increase of the distal involved lung with secondary hypoplasia of the normal lung, and 2) peripheral, located at the segmental/subsegmental bronchial level, which may present as an isolated lesion or as part of a complex congenital malformation. Prenatal findings are mostly nonspecific. Postnatal exams show overinflated lung areas and focal bronchial dilations. The typical fluid-filled bronchoceles are not always observed in neonates but develop progressively in the first months of life. This pictorial essay describes the spectrum of imaging findings of bronchial atresia in fetuses, neonates and infants.

Determination of fetal lung maturity using magnetic resonance imaging signal intensity measurements

PURPOSE

The objective of this study was to determine if magnetic resonance signal intensity measurements can be used to predict gestational age and hence fetal lung maturity.

METHODS

This institutional review board-approved study was a retrospective review of 394 fetal magnetic resonance imaging cases from a single institution for the years 2001 to 2011. For each case, T1- and T2-weighted sequences were selected for data collection. A single reviewer obtained 10 regions of interest (when possible) from each scan (fetal lung, fetal liver, fetal muscle, fetal spleen, and maternal urine, for both T1- and T2-weighted sequences). The medical record was searched for relevant information including best estimate of gestational age, Apgar scores, karyotype, and fetal diagnosis. A variety of organ-to-organ ratios and direct organ signal intensity measurements were assessed for correlation with gestational age.

RESULTS

Three hundred thirty-five cases met inclusion criteria with gestational ages ranging from 17 to 39 weeks (mean, 28.6 weeks). A significant relationship between magnetic resonance signal intensity ratios and gestational age was demonstrated on the T2 lung-to-liver, T2 lung-to-spleen, T2 lung-to-muscle, T1 lung-to-liver, and T1 lung-to-spleen ratios (P < 0.05). T2 lung-to-liver and T2 lung-to-muscle demonstrated the strongest relationship with gestational age (best correlation r = 0.483, P < 0.001). T1 lung-to-liver and T1 lung-to-spleen demonstrated inverse relationships with gestational age (r = -0.174 [P = 0.03] and r = -0.236 [P = 0.02], respectively).

CONCLUSIONS

A significant correlation between multiple signal intensity ratios and gestational age is demonstrated. However, the large variances preclude a clinically useful relationship.

Prenatal diagnosis of fetal respiratory function: evaluation of fetal lung maturity using lung-to-liver signal intensity ratio at magnetic resonance imaging

Objective

The purpose of this retrospective study is to determine the fetal lung-to-liver signal intensity ratio (LLSIR) on T2-weighted images for the prediction of neonatal respiratory outcome.

Methods

One hundred ten fetuses who underwent magnetic resonance imaging (MRI) examination for various indications after 22 weeks of gestation participated in this study. LLSIR was measured as the ratio of signal intensities of the fetal lung and liver on T2-weighted images at MRI. We examined the changes of the ratio with advancing gestation and the relations between LLSIR and the presence of the severe respiratory disorder (SRD) after birth. The best cut-off value of the LLSIR to predict respiratory outcome after birth was calculated using receiver operating characteristic (ROC) curve analysis.

Results

Lung-to-liver signal intensity ratio correlated significantly with advancing gestational age (R = 0.35, p < 0.001). The non-SRD group had higher LLSIR compared with the SRD group (2.15 ± 0.30 vs. 1.53 ± 0.40, p < 0.001). ROC curve analysis showed that fetuses with an LLSIR < 2.00 were more likely to develop SRD [sensitivity: 100%, 95% confidence interval (CI): 52–100%; specificity: 73%, 95% CI 54–88%].

Conclusion

The fetal LLSIR on T2-weighted images is an accurate marker to diagnose the fetal lung maturity. © 2014 The Authors. Prenatal Diagnosis published by John Wiley & Sons, Ltd.

Standardized fetal anatomical examination using magnetic resonance imaging: a feasibility study

OBJECTIVE

To determine whether a standard complete fetal anatomical survey, as recommended for ultrasound examination guidelines, is feasible using a standardized magnetic resonance imaging (MRI) protocol.

METHODS

Based on guidelines for ultrasound examination, we created a specific MRI protocol for fetal anatomical survey. This protocol was then tested prospectively in 100 women undergoing fetal MRI examination for various specific indications at a median gestational age of 30 weeks. The feasibility of using MRI to perform the fetal anatomical survey was analyzed by two reviewers (A and B) based on 26 predefined anatomical criteria, yielding a score ranging from 0 to 26 (26 meaning successful complete anatomical study). Reproducibility was analyzed using percentage agreement and modified kappa statistics.

RESULTS

The mean score for the standardized MRI anatomical survey was 24.6 (SD, 1.4; range, 15-26) for Reviewer A and 24.2 (SD, 1.7; range, 15-26) for Reviewer B (P = 0.1). Twenty-two, two and two criteria could be assessed in > 95%, 80-95% and < 80% of cases by Reviewer A and 19, four and three criteria could be assessed in > 95%, 80-95% and < 80% of cases by Reviewer B. For both reviewers, the two most difficult criteria to evaluate were aorta and pulmonary artery. Inter-reviewer agreement was above 90% for 22 of the 26 anatomical criteria and adjusted kappa coefficients for each criterion demonstrated good, moderate and poor agreement for 22, two and two criteria, respectively.

CONCLUSION

Our data support the hypothesis that standardized fetal anatomical examination might be achieved and reproducible using MRI, although improvement is required for the cardiac part of the examination.

Maternal smoking and fetal lung signals--an in utero MRI investigation

OBJECTIVE

To investigate whether fetal lung signals and fetal lung signal progression over gestation observed on magnetic resonance imaging are different in mothers who reported smoking during pregnancy compared with nonsmoking controls.

METHOD

Cross-sectional retrospective study of 100 consecutive singleton pregnancies that underwent magnetic resonance imaging. Fetal lung-liver signal intensity ratios of 18 fetuses of mothers who reported smoking during pregnancy were compared with 82 fetuses of nonsmoking controls.

RESULTS

Average gestational age at magnetic resonance imaging was 26.4 ± 5.2 weeks (Range 18.4-38.2 weeks). Cases reported smoking between 2 and 15 cigarettes per day. The mean number of cigarettes per day for cases was 9.2 ± 3.4. Mean fetal lung-liver signal intensity ratios did not differ significantly between the two groups (p = 0.8). They showed a linear increase with gestational age (r(2) = 0.3). Multiple regression analysis of lung-liver signal intensity ratios using gestational age and smoking status as predictors revealed a significant influence of gestational age (p < 0.0001) but not maternal smoking status (p = 0.8) on fetal lung-liver signal intensity ratios.

CONCLUSIONS

Fetuses of mothers who reported smoking during pregnancy show similar lung signals and lung signal progression over gestation on magnetic resonance imaging as nonsmoking controls.

Abnormalities of the upper extremities on fetal magnetic resonance imaging

OBJECTIVE

In view of the increasing use of fetal magnetic resonance imaging (MRI) as an adjunct to prenatal ultrasonography, we sought to demonstrate the visualization of upper extremity abnormalities and associated defects on MRI, with regard to fetal outcomes and compared with ultrasound imaging.

METHODS

This retrospective study included 29 fetuses with upper extremity abnormalities visualized with fetal MRI following suspicious ultrasound findings and confirmed by postnatal assessment or autopsy. On a 1.5-Tesla unit, dedicated sequences were applied to image the extremities. Central nervous system (CNS) and extra-CNS anomalies were assessed to define extremity abnormalities as isolated or as complex, with associated defects. Fetal outcome was identified from medical records. MRI and ultrasound findings, when available, were compared.

RESULTS

Isolated upper extremity abnormalities were found in three (10.3%) fetuses. In 26 (89.7%) fetuses complex abnormalities, including postural extremity disorders (21/26) and structural extremity abnormalities (15/26), were demonstrated. Associated defects involved: face (15/26); musculoskeletal system (14/26); thorax and cardio/pulmonary system (12/26); lower extremities (12/26); brain and skull (10/26); and abdomen (8/26). Of the 29 cases, 18 (62.1%) pregnancies were delivered and 11 (37.9%) were terminated. MRI and US findings were compared in 27/29 cases: the diagnosis was concordant in 14 (51.9%) of these cases, and additional findings were made on MRI in 13/27 (48.1%) cases.

CONCLUSIONS

Visualization of upper extremity abnormalities on fetal MRI enables differentiation between isolated defects and complex ones, which may be related to poor fetal prognosis. MRI generally confirms the ultrasound diagnosis, and may provide additional findings in certain cases.

Effect of antenatal corticosteroid treatment on the fetal lung: a magnetic resonance imaging study

OBJECTIVES

To investigate the effect of antenatal corticosteroid treatment on the fetal lung using magnetic resonance imaging.

METHODS

Prospective evaluation of 30 consecutive singleton pregnancies that received antenatal corticosteroid treatment (12 mg betamethasone i.m. on admission and 24 h later) because of threatened preterm birth. Fetal lungs were assessed using T2-weighted single-shot fast spin-echo images of a whole-body 1.5-T superconducting unit twice: less than 24 h and more than 48 h after the first course of betamethasone. Lung volumes and lung-liver signal-intensity ratios were compared between the two time points.

RESULTS

Nine patients had to be excluded from the analysis because they did not complete the study protocol as required. Ten female and 11 male fetuses with a gestational age between 23.4 and 32.6 weeks were included in the final analysis. The mean gestational age of included fetuses was 27.5 ± 2.8 weeks. Using a linear regression model, a significant influence of gestational age on ln fetal lung volume (r(2)=0.414; P<0.0001) and lung-liver signal-intensity ratios (r(2)=0.271, P<0.0001) was found. Between the two evaluated time points, a significant increase in lung-liver signal-intensity ratios (2.34 ± 0.72 vs. 3.22 ± 1.12, P<0.0001), but not in mean lung volumes (46.6 ± 20.7 cm(3) vs. 48.8 ± 16.0 cm(3) , P=0.292), was observed.

CONCLUSION

We demonstrate a significant increase in lung-liver signal-intensity ratios after antenatal corticosteroid treatment for induction of lung maturation which most likely reflects changing properties of the fetal lung parenchyma. This could potentially be useful in non-invasively assessing the effect of antenatal corticosteroid treatment on the lungs of fetuses at risk for preterm birth.

Maternal smoking and fetal lung volume-an in utero MRI investigation

OBJECTIVE

To investigate whether fetal lung volume and fetal lung volume growth over gestation are different in mothers who reported smoking during pregnancy compared to non-smoking controls.

METHOD

Cross-sectional retrospective study of the data of 200 consecutive singleton pregnancies that underwent magnetic resonance imaging (MRI). Fetal lung volumes of 32 fetuses of mothers who reported smoking during pregnancy were compared to 168 fetuses of non-smoking controls.

RESULTS

Cases reported smoking between 2 and 30 cigarettes per day. The mean number of cigarettes per day for cases was 10.2 ± 6.1. After logarithmic transformation, lung volumes showed a linear increase with gestational age (r(2) = 0.7). Multiple regression analysis revealed a significant influence of gestational age (p < 0.0001) but not maternal smoking status (p = 1.0) on fetal lung volumes. Comparison of the relative difference between observed and predicted lung volumes for both groups by means of unpaired t-test revealed no significant influence of maternal smoking status on this variable (p = 0.9).

CONCLUSIONS

Fetuses of mothers who reported smoking during pregnancy show similar lung volumes and lung volume growth in MRI compared to non-smoking controls. Our data suggest that a reduced lung volume is not responsible for the increase in respiratory symptoms associated with prenatal smoking.

Prenatal ultrasound and fetal MRI: the comparative value of each modality in prenatal diagnosis

Fetal MRI is used with increasing frequency as an adjunct to ultrasound (US) in prenatal diagnosis. In this review, we discuss the relative value of both prenatal US and MRI in evaluating fetal and extra-fetal structures for a variety of clinical indications. Advantages and disadvantages of each imaging modality are addressed. In summary, MRI has advantages in demonstrating pathology of the brain, lungs, complex syndromes, and conditions associated with reduction of amniotic fluid. At present, US is the imaging method of choice during the first trimester, and in the diagnosis of cardiovascular abnormalities, as well as for screening. In some conditions, such as late gestational age, increased maternal body mass index, skeletal dysplasia, and metabolic disease, neither imaging method may provide sufficient diagnostic information.

MRI of the lungs in children

Lung diseases of children often need diagnostic imaging beyond X-ray. Although CT is considered the gold standard of lung imaging, MRI is sufficient to answer most of the questions raised. After all, the exposure to radiation caused by one CT examination corresponds to approximately the effective dose of 200 chest radiographs. What is MRI's potential in the lung today? In diseases with alveolar pathology, cardiac- and respiratory-triggered MRI examinations are roughly equivalent to CT examinations. Distinct interstitial processes are easily diagnosable using MRI. Early interstitial processes may be missed by MRI, but conventional plain films fail to recognize them just as often. For identification of lung metastases, CT is still used as the initial diagnostic measure. Subsequent therapy monitoring may then be carried out with the help of MRI. Small bullae and pulmonary emphysema at present pose a problem to MRI. On the other hand, MRI is reliable for follow-up examinations in inflammatory diseases or for imaging of complications, and the increased use of lung MRI as an alternative to chest CT may contribute immensely to reducing radiation exposure in children.

[Fetal lung imaging]

Exponential improvements in imaging techniques over the last ten years, through patients' and physicians' wishes for less invasive fetal work-up, now allow us to better explore and understand fetal lung physiology during pregnancy. Diagnostic and prognostic consequences at stake are huge, especially for fetuses at risk of pulmonary hypoplasia. We will decline in three parts (normal lung, malformative lung and pulmonary hypoplasia), through a review of the literature and at the light of our experience, the potentialities and limitations of all imaging modalities (Ultrasound, Doppler, 3D, MRI). Then, we will dwell on future leads and the need for large-scale collaborative studies.