Contrast-enhanced 4D-MRI for internal target volume generation in treatment planning for liver tumors.

BACKGROUND: Liver tumors are often invisible on four-dimensional commuted tomography (4D-CT). Imperfect imaging surrogates are used to estimate the tumor motion. Here, we assessed multiple 4D magnetic resonance (MR) binning algorithms for directly visualizing liver tumor motion for radiotherapy planning. METHODS: Patients were simulated using a 3 Tesla MR and CT scanner. Three prototype binning algorithms (phase, amplitude, and two-directional) were applied to the 4D-MRIs, and the image quality was assessed using a qualitative clarity score and quantitative sharpness score. Radiation plans were generated for internal target volumes (ITVs) derived using 4D-MRI and 4D-CT, and the dosimetry of targets were compared. Paired t-tests were used to compare sharpness scores and dosimetric data. RESULTS: Twelve patients with 17 liver tumors were scanned between May and November 2021. Compared to phase binning, two-directional demonstrated equal or better clarity and sharpness scores (end-expiration: 0.33 vs 0.38, p = 0.018, end-inspiration: 0.28 vs 0.31, p = 0.010). Compared to amplitude binning, two-directional binning captured hysteresis of ≥ 3 mm in 35 % of patients. Evaluation of dosimetry CT-optimized plans revealed that PTV coverage of MR-derived targets were significantly lower than CT-derived targets (PTV receiving 90 % of prescription: 75.56 % vs 89.38 %, p = 0.002). CONCLUSION: Using contrast-enhanced 4D-MRI is feasible for directly delineating liver tumors throughout the respiratory cycle. The current standard of using radiation plans optimized for 4D-CT-derived targets achieved lower coverage of directly visualized MRI targets, suggesting that adopting MRI for motion management may improve radiation treatment of liver lesions and reduce the risk of marginal misses.

MR-assisted PET respiratory motion correction using deep-learning based short-scan motion fields.

PURPOSE: We evaluated the impact of PET respiratory motion correction (MoCo) in a phantom and patients. Moreover, we proposed and examined a PET MoCo approach using motion vector fields (MVFs) from a deep-learning reconstructed short MRI scan. METHODS: The evaluation of PET MoCo was performed in a respiratory motion phantom study with varying lesion sizes and tumor to background ratios (TBRs) using a static scan as the ground truth. MRI-based MVFs were derived from either 2000 spokes (MoCo2000 , 5-6 min acquisition time) using a Fourier transform reconstruction or 200 spokes (MoCoP2P200 , 30-40 s acquisition time) using a deep-learning Phase2Phase (P2P) reconstruction and then incorporated into PET MoCo reconstruction. For six patients with hepatic lesions, the performance of PET MoCo was evaluated using quantitative metrics (SUVmax , SUVpeak , SUVmean , lesion volume) and a blinded radiological review on lesion conspicuity. RESULTS: MRI-assisted PET MoCo methods provided similar results to static scans across most lesions with varying TBRs in the phantom. Both MoCo2000 and MoCoP2P200 PET images had significantly higher SUVmax , SUVpeak , SUVmean and significantly lower lesion volume than non-motion-corrected (non-MoCo) PET images. There was no statistical difference between MoCo2000 and MoCoP2P200 PET images for SUVmax , SUVpeak , SUVmean or lesion volume. Both radiological reviewers found that MoCo2000 and MoCoP2P200 PET significantly improved lesion conspicuity. CONCLUSION: An MRI-assisted PET MoCo method was evaluated using the ground truth in a phantom study. In patients with hepatic lesions, PET MoCo images improved quantitative and qualitative metrics based on only 30-40 s of MRI motion modeling data.

Phase2Phase: Respiratory Motion-Resolved Reconstruction of Free-Breathing Magnetic Resonance Imaging Using Deep Learning Without a Ground Truth for Improved Liver Imaging.

OBJECTIVES: Respiratory binning of free-breathing magnetic resonance imaging data reduces motion blurring; however, it exacerbates noise and introduces severe artifacts due to undersampling. Deep neural networks can remove artifacts and noise but usually require high-quality ground truth images for training. This study aimed to develop a network that can be trained without this requirement. MATERIALS AND METHODS: This retrospective study was conducted on 33 participants enrolled between November 2016 and June 2019. Free-breathing magnetic resonance imaging was performed using a radial acquisition. Self-navigation was used to bin the k-space data into 10 respiratory phases. To simulate short acquisitions, subsets of radial spokes were used in reconstructing images with multicoil nonuniform fast Fourier transform (MCNUFFT), compressed sensing (CS), and 2 deep learning methods: UNet3DPhase and Phase2Phase (P2P). UNet3DPhase was trained using a high-quality ground truth, whereas P2P was trained using noisy images with streaking artifacts. Two radiologists blinded to the reconstruction methods independently reviewed the sharpness, contrast, and artifact-freeness of the end-expiration images reconstructed from data collected at 16% of the Nyquist sampling rate. The generalized estimating equation method was used for statistical comparison. Motion vector fields were derived to examine the respiratory motion range of 4-dimensional images reconstructed using different methods. RESULTS: A total of 15 healthy participants and 18 patients with hepatic malignancy (50 ± 15 years, 6 women) were enrolled. Both reviewers found that the UNet3DPhase and P2P images had higher contrast (P < 0.01) and fewer artifacts (P < 0.01) than the CS images. The UNet3DPhase and P2P images were reported to be sharper than the CS images by 1 reviewer (P < 0.01) but not by the other reviewer (P = 0.22, P = 0.18). UNet3DPhase and P2P were similar in sharpness and contrast, whereas UNet3DPhase had fewer artifacts (P < 0.01). The motion vector lengths for the MCNUFFT800 and P2P800 images were comparable (10.5 ± 4.2 mm and 9.9 ± 4.0 mm, respectively), whereas both were significantly larger than CS2000 (7.0 ± 3.9 mm; P < 0.0001) and UNnet3DPhase800 (6.9 ± 3.2; P < 0.0001) images. CONCLUSIONS: Without a ground truth, P2P can reconstruct sharp, artifact-free, and high-contrast respiratory motion-resolved images from highly undersampled data. Unlike the CS and UNet3DPhase methods, P2P did not artificially reduce the respiratory motion range.

Dual-Imaging Modality Approach to Evaluate Cerebral Hemodynamics in Growth-Restricted Fetuses: Oxygenation and Perfusion.

OBJECTIVE: To evaluate a dual-imaging modality approach to obtain a combined estimation of venous blood oxygenation (SνO2) using susceptibility-weighted magnetic resonance imaging (SWI-MRI), and blood perfusion using power Dopp-ler ultrasound (PDU) and fractional moving blood volume (FMBV) in the brain of normal growth and growth-restricted fetuses. METHODS: Normal growth (n = 33) and growth-restricted fetuses (n = 10) from singleton pregnancies between 20 and 40 weeks of gestation were evaluated. MRI was performed and SνO2 was calculated using SWI-MRI data obtained in the straight section of the superior sagittal sinus. Blood perfusion was estimated using PDU and FMBV from the frontal lobe in a mid-sagittal plane of the fetal brain. The association between fetal brain SνO2 and FMBV, and the distribution of SνO2 and FMBV values across gestation were calculated for both groups. RESULTS: In growth-restricted fetuses, the brain SνO2 values were similar, and the FMBV values were higher across gestation as compared to normal growth fetuses. There was a significantly positive association between SνO2 and FMBV values (slope = 0.38 ± 0.12; r = 0.7; p = 0.02) in growth-restricted fetuses. In normal growth fetuses, SνO2 showed a mild decreasing trend (slope = -0.7 ± 0.4; p = 0.1), whereas FMBV showed a mild increasing trend (slope = 0.2 ± 0.2; p = 0.2) with advancing gestation, and a mild but significant negative association (slope = -0.78 ± 0.3; r = -0.4; p = 0.04) between these two estimates. CONCLUSION: Combined MRI (SWI) and ultrasound (FMBV) techniques showed a significant association between cerebral blood oxygenation and blood perfusion in normal growth and growth-restricted fetuses. This dual-imaging approach could contribute to the early detection of fetal "brain sparing" and brain oxygen saturation changes in high-risk pregnancies.

Quantitative susceptibility mapping in the human fetus to measure blood oxygenation in the superior sagittal sinus.

OBJECTIVES: To present the feasibility of performing quantitative susceptibility mapping (QSM) in the human fetus to evaluate the oxygenation (SvO2) of cerebral venous blood in vivo. METHODS: Susceptibility weighted imaging (SWI) data were acquired from healthy pregnant subjects (n = 21, median = 31.3 weeks, interquartile range = 8.8 weeks). The susceptibility maps were generated from the SWI-phase images using a modified QSM processing pipeline, optimised for fetal applications. The processing pipeline is as follows: (1) mild high-pass filtering followed by quadratic fitting of the phase images to eliminate background phase variations; (2) manual creation of a fetal brain mask that includes the superior sagittal sinus (SSS); (3) inverse filtering of the resultant masked phase images using a truncated k-space approach with geometric constraint. Further, the magnetic susceptibility, ∆χv and corresponding putative SvO2 of the SSS were quantified from the generated susceptibility maps. Systematic error in the measured SvO2 due to the modified pipeline was also studied through simulations. RESULTS: Simulations showed that the systematic error in SvO2 when using a mask that includes a minimum of 5 voxels around the SSS and five slices remains < 3% for different orientations of the vessel relative to the main magnetic field. The average ∆χv in the SSS quantified across all gestations was 0.42 ± 0.03 ppm. Based on ∆χv, the average putative SvO2 in the SSS across all fetuses was 67% ± 7%, which is in good agreement with published studies. CONCLUSIONS: This in vivo study demonstrates the feasibility of using QSM in the human fetal brain to estimate ∆χv and SvO2. KEY POINTS: • A modified quantitative susceptibility mapping (QSM) processing pipeline is tested and presented for the human fetus. • QSM is feasible in the human fetus for measuring magnetic susceptibility and oxygenation of venous blood in vivo. • Blood magnetic susceptibility values from MR susceptometry and QSM agree with each other in the human fetus.

Quantitative Flow Imaging in Human Umbilical Vessels In Utero Using Nongated 2D Phase Contrast MRI.

BACKGROUND: Volumetric assessment of afferent blood flow rate provides a measure of global organ perfusion. Phase-contrast magnetic resonance imaging (PCMRI) is a reliable tool for volumetric flow quantification, but given the challenges with motion and lack of physiologic gating signal, such studies, in vivo on the human placenta, are scant. PURPOSE: To evaluate and apply a nongated (ng) PCMRI technique for quantifying blood flow rates in utero in umbilical vessels. STUDY TYPE: Prospective study design. STUDY POPULATION: Twenty-four pregnant women with median gestational age (GA) 30 4/7 weeks and interquartile range (IQR) 8 1/7 weeks. FIELD STRENGTH/SEQUENCE: All scans were performed on a 3.0T Siemens Verio system using the ng-PCMRI technique. ASSESSMENT: The GA-dependent increase in umbilical vein (UV) and arterial (UA) flow was compared to previously published values. Systematic error to be expected from ng-PCMRI, in the context of pulsatile UA flow and partial voluming, was studied through Monte-Carlo simulations, as a function of resolution and number of averages. STATISTICAL TESTS: Correlation between the UA and UV was evaluated using a generalized linear model. RESULTS: Simulations showed that ng-PCMRI measurement variance reduced by increasing the number of averages. For vessels on the order of 2 voxels in radius, partial voluming led to 10% underestimation in the flow. In fetuses, the average flow rates in UAs and UV were measured to be 203 ± 80 ml/min and 232 ± 92 ml/min and the normalized average flow rates were 140 ± 59 ml/min/kg and 155 ± 57 ml/min/kg, respectively. Excellent correlation was found between the total arterial flow vs. corresponding venous flow, with a slope of 1.08 (P = 0.036). DATA CONCLUSION: Ng-PCMRI can provide accurate volumetric flow measurements in utero in the human umbilical vessels. Care needs to be taken to ensure sufficiently high-resolution data are acquired to minimize partial voluming-related errors. LEVEL OF EVIDENCE: 2 Technical Efficacy Stage 1 J. Magn. Reson. Imaging 2017.

Imaging putative foetal cerebral blood oxygenation using susceptibility weighted imaging (SWI).

OBJECTIVE: To evaluate the magnetic susceptibility, ∆χ v , as a surrogate marker of venous blood oxygen saturation, S v O 2, in second- and third-trimester normal human foetuses. METHODS: Thirty-six pregnant women, having a mean gestational age (GA) of 31 2/7 weeks, underwent magnetic resonance imaging (MRI). Susceptibility-weighted imaging (SWI) data from the foetal brain were acquired. ∆χ v of the superior sagittal sinus (SSS) was quantified using MR susceptometry from the intra-vascular phase measurements. Assuming the magnetic property of foetal blood, ∆χ do , is the same as that of adult blood, S v O 2 was derived from the measured Δχ v . The variation of ∆χ v and S v O 2, as a function of GA, was statistically evaluated. RESULTS: The mean ∆χ v in the SSS in the second-trimester (n = 8) and third-trimester foetuses (n = 28) was found to be 0.34± 0.06 ppm and 0.49 ±0.05 ppm, respectively. Correspondingly, the derived S v O 2 values were 69.4% ±3.27% and 62.6% ±3.25%. Although not statistically significant, an increasing trend (p = 0.08) in Δχ v and a decreasing trend (p = 0.22) in S v O 2 with respect to advancing gestation was observed. CONCLUSION: We report cerebral venous blood magnetic susceptibility and putative oxygen saturation in healthy human foetuses. Cerebral oxygen saturation in healthy human foetuses, despite a slight decreasing trend, does not change significantly with advancing gestation. KEY POINTS: • Cerebral venous magnetic susceptibility and oxygenation in human foetuses can be quantified. • Cerebral venous oxygenation was not different between second- and third-trimester foetuses. • Foetal cerebral venous oxygenation does not change significantly with advancing gestation.

Magnetic resonance angiography of fetal vasculature at 3.0 T.

Magnetic resonance angiography has not been used much previously for visualizing fetal vessels in utero for reasons that include a contraindication for the use of exogenous contrast agents, maternal respiratory motion and fetal motion. In this work, we report the feasibility of using an appropriately modified clinical time-of-flight magnetic resonance imaging sequence for non-contrast angiography of human fetal and placental vessels at 3.0 T. Using this 2D angiography technique, it is possible to visualize fetal vascular networks in late pregnancy. KEY POINTS: • 3D-visualization of fetal vasculature is feasible using non-contrast MRA at 3.0 T. • Visualization of placental vasculature is also possible with this method. • Fetal MRA can serve as a vascular localizer for quantitative MRI studies. • This method can be extended to 1.5 T.

A longitudinal study of placental perfusion using dynamic contrast enhanced magnetic resonance imaging in murine pregnancy.

INTRODUCTION: To evaluate changes in placental perfusion with advancing gestation in normal murine pregnancy using dynamic contrast enhanced magnetic resonance imaging (DCE-MRI). METHODS: Seven timed-pregnant CD-1 mice underwent DCE-MRI scanning longitudinally on gestational days (GD) 13, 15 and 17. Placentas were segmented into high (HPZ) and low perfusion zones (LPZ) using tissue similarity mapping. Blood perfusion of the respective regions and the whole placenta was quantified using the steepest slope method. The diameter of the maternal central canal (CC) was also measured. RESULTS: An increase in perfusion was observed between GD13 and GD17 in the overall placenta (p = 0.04) and in the HPZ (p = 0.02). Although perfusion in the LPZ showed a slight increasing trend, it was not significant (p = 0.07). Perfusion, in units of ml/min/100 ml, in the overall placenta and the HPZ was respectively 61.2 ± 31.2 and 106.2 ± 56.3 at GD13 (n = 19 placentas); 90.3 ± 43.7 and 139 ± 55.4 at GD15 (n = 20); and 104.9 ± 76.1 and 172.2 ± 85.6 at GD17 (n = 14). The size of the CC increased with advancing gestation (p < 0.05). DISCUSSION: Using longitudinal DCE-MRI, the gestational age-dependent perfusion change in the normal murine placenta and in its regional compartments was quantified. In mid and late gestations, placental constituent regions differ significantly in their perfusion rates. The CC diameter also showed increase with advancing gestation, which may be playing an important role toward the gestational age-dependent increase in placental perfusion.

Longitudinal Changes in Placental Magnetic Resonance Imaging Relaxation Parameter in Murine Pregnancy: Compartmental Analysis.

OBJECTIVE: To quantify gestation-dependent longitudinal changes in the magnetic resonance transverse relaxation time (T2) parameter of the major constituent regions of the mouse placenta and to evaluate their relative contributions to changes in overall placental T2. METHODS: Timed-pregnant CD-1 mice underwent magnetic resonance imaging at 7.0 T field strength, on gestational day 13 (GD13), GD15 and GD17. T2 of the placenta and its constituent high and low blood perfusion regions were quantified. A linear mixed-effects model was used to fit the T2 across gestation, and the significance of coefficients was tested. RESULTS: A decrease in the T2 values of the placenta and its constituent regions was observed across gestation. The temporal change in T2 was estimated to be -1.85 ms/GD (p < 0.0001) for the placenta, -1.00 ms/GD (p < 0.001) for the high-perfusion zones (HPZs) and -1.66 ms/GD (p < 0.0001) for the low-perfusion zones (LPZs). CONCLUSION: T2 of the constituent zones of the murine placenta decreases with advancing gestation. While the T2 of the LPZ is smaller than that of the HPZ, there is no difference in their decrease rate relative to that of the whole placenta (p = 0.24). The results suggest an increased role of constituent volume fractions in affecting overall gestation-dependent placental T2 decrease in mice.

MR imaging of the fetal brain at 1.5T and 3.0T field strengths: comparing specific absorption rate (SAR) and image quality.

OBJECTIVES: Our two objectives were to evaluate the feasibility of fetal brain magnetic resonance imaging (MRI) using a fast spin echo sequence at 3.0T field strength with low radio frequency (rf) energy deposition (as measured by specific absorption rate: SAR) and to compare image quality, tissue contrast and conspicuity between 1.5T and 3.0T MRI. METHODS: T2 weighted images of the fetal brain at 1.5T were compared to similar data obtained in the same fetus using a modified sequence at 3.0T. Quantitative whole-body SAR and normalized image signal to noise ratio (SNR), a nominal scoring scheme based evaluation of diagnostic image quality, and tissue contrast and conspicuity for specific anatomical structures in the brain were compared between 1.5T and 3.0T. RESULTS: Twelve pregnant women underwent both 1.5T and 3.0T MRI examinations. The image SNR was significantly higher (P=0.03) and whole-body SAR was significantly lower (P<0.0001) for images obtained at 3.0T compared to 1.5T. All cases at both field strengths were scored as having diagnostic image quality. Images from 3.0T MRI (compared to 1.5T) were equal (57%; 21/37) or superior (35%; 13/37) for tissue contrast and equal (61%; 20/33) or superior (33%, 11/33) for conspicuity. CONCLUSIONS: It is possible to obtain fetal brain images with higher resolution and better SNR at 3.0T with simultaneous reduction in SAR compared to 1.5T. Images of the fetal brain obtained at 3.0T demonstrated superior tissue contrast and conspicuity compared to 1.5T.

Quantitative T2 changes and susceptibility-weighted magnetic resonance imaging in murine pregnancy.

OBJECTIVE: To evaluate gestational age-dependent changes in the T2 relaxation time in normal murine placentas in vivo. The role of susceptibility-weighted imaging (SWI) in visualization of the murine fetal anatomy was also elucidated. METHODS: Timed-pregnant CD-1 mice at gestational day (GD) 12 and GD17 underwent magnetic resonance imaging. Multi-echo spin echo and SWI data were acquired. The placental T2 values on GD12 and GD17 were quantified. To account for the influence of systemic maternal physiological factors on placental perfusion, maternal muscle was used as a reference for T2 normalization. A linear mixed-effects model was used to fit the normalized T2 values, and the significance of the coefficients was tested. Fetal SWI images were processed and reviewed for venous vasculature and skeletal structures. RESULTS: The average placental T2 value decreased significantly on GD17 (40.17 ± 4.10 ms) compared to the value on GD12 (55.78 ± 8.13 ms). The difference in normalized T2 values also remained significant (p = 0.001). Using SWI, major fetal venous structures like the cardinal vein, the subcardinal vein, and the portal vein were visualized on GD12. In addition, fetal skeletal structures could also be discerned on GD17. CONCLUSION: The T2 value of a normal murine placenta decreases with advancing gestation. SWI provided clear visualization of the fetal venous vasculature and bony structures. © 2014 S. Karger AG, Basel.

Measuring venous blood oxygenation in fetal brain using susceptibility-weighted imaging.

PURPOSE: To evaluate fetal cerebral venous blood oxygenation, Yv, using principles of MR susceptometry. MATERIALS AND METHODS: A cohort of 19 pregnant subjects, with a mean gestational age of 31.6 ± 4.7 weeks were imaged using a modified susceptibility-weighted imaging (SWI) sequence. Data quality was first assessed for feasibility of oxygen saturation measurement, and data from five subjects (mean ± std gestational age of 33.7 ± 3.6 weeks) were then chosen for further quantitative analysis. SWI phase in the superior sagittal sinus was used to evaluate oxygen saturation using the principles of MR susceptometry. Systematic error in the measured Y(v) values was studied through simulations. RESULTS: Simulations showed that the systematic error in Yv depended upon the assumed angle of the vessel, θ, relative to the main magnetic field and the error in that vessel angle δθ. For the typical vessel angle of θ = 30° encountered in the fetal data analyzed, a δθ as large as ±20° led to an absolute error, δYv, of less than 11%. The measured mean oxygen saturation across the five fetuses was 66% ± 9.4%. This average cerebral venous blood oxygenation value is in close agreement with values in the published literature. CONCLUSION: We have reported the first in vivo measurement of human fetal cerebral venous oxygen saturation using MRI.