Motion corrected fetal body MRI provides reliable 3D lung volumes in normal and abnormal fetuses

Objective:Evaluate deformable slice-to-volume registration (DSVR) to calculate 3D-segmented total lung volume (TLV) in fetuses with congenital diaphragmatic hernia, congenital lung lesions and healthy controls, with comparison to 2D-manual segmentation. Design:Pilot study Setting:Regional fetal medicine referral centre Sample:Fetal MRIs performed for clinical indications (abnormal cases) or as research participants (healthy controls) Methods:Sixteen MRI datasets of fetuses (22-32 weeks GA). Diagnosis: CDH(n=5), CPAM(n=2), CDH with BPS(n=1) and healthy control(n=8). DSVR was used for reconstruction of 3D isotropic (0.85 mm) volumes of fetal body followed by semi-automated lung segmentation. The resulting 3D TLV were compared to the traditional 2D-based volumetry, and a normogram of DSVR-derived fetal lung volumes from 100 cases was produced. Main Outcome Measures:Concordance with 2D-volumetry assessed with Bland-Altman analysis, results of segmentations presented visually. Observed/Expected values were calculated for abnormal cases based upon the normogram. Results:DSVR-derived TLV values have high correlation with the 2D-based measurements but with a consistently lower volume; bias -1.44cm3 [95% limits: -2.6 to -0.3] with improved resolution able to exclude hilar structures even in severe motion corruption or in cases of lung hypoplasia. Conclusions:Application of DSVR for fetal MRI provides a solution for analysis of motion corrupted scans and does not suffer from the interpolation error inherent in 2D-segmentation as per current clinical practice. It increases information content of acquired data in terms of visualising organs in 3D space and quantification of volumes, which we believe will have important value for counselling and surgical planning. Keywords:Fetal MRI; congenital diaphragmatic hernia; CPAM; lung volume

Automated 3D reconstruction of the fetal thorax in the standard atlas space from motion-corrupted MRI stacks for 21-36 weeks GA range

Slice-to-volume registration (SVR) methods allow reconstruction of high-resolution 3D images from multiple motion-corrupted stacks. SVR-based pipelines have been increas- ingly used for motion correction for fetal MRI since they allow more informed and de- tailed diagnosis of brain and body anomalies including congenital heart defects (Lloyd et al., 2019). Recently, fully automated rigid SVR reconstruction of the fetal brain in the atlas space was achieved in (Salehi et al., 2019) that used segmentation and pose es- timation convolutional neural networks (CNNs). However, these CNN-based methods have not yet been applied to the fetal body region. Meanwhile, the existing rigid and deformable SVR (DSVR) solutions (Uus et al., 2020) for the fetal trunk region are limited by the requirement of manual input as well the narrow capture range of the classical gradient descent based registration methods that cannot resolve severe fetal motion fre- quently occurring at the early gestational age (GA). Furthermore, in our experience, the conventional 2D slice-wise CNN-based brain masking solutions are reportedly prone to errors that require manual corrections when applied on a wide range of acquisition protocols or abnormal cases in clinical setting. In this work, we propose a fully automated pipeline for reconstruction of the fetal thorax region for 21-36 weeks GA range MRI datasets. It includes 3D CNN-based intra-uterine localisation of the fetal trunk and landmark-guided pose estimation steps that allow automated DSVR reconstruction in the standard radiological space irrespec- tive of the fetal body position or the regional stack coverage. The additional step for generation of the common template space and rejection of outliers provides the means for automated exclusion of stacks affected by low image quality or extreme motion. The pipeline was evaluated on a series of experiments including fetal MRI datasets and simulated rotation motion. Furthermore, we performed a qualitative assessment of the image reconstruction quality in terms of the definition of vascular structures on 100 early (median 23.14 weeks) and late (median 31.79 weeks) GA group MRI datasets covering 21 to 36 weeks GA range.

Spatiotemporal Changes in Diffusivity and Anisotropy in Fetal Brain Tractography

Population averaged diffusion atlases can be utilized to characterize complex microstructural changes with less bias than data from individual subjects. In this study, a fetal diffusion tensor imaging (DTI) atlas was used to investigate tract-based changes in anisotropy and diffusivity in vivo from 23 to 38 weeks of gestational age (GA). Healthy pregnant volunteers with typically developing fetuses were imaged at 3 Tesla. Acquisition included structural images processed with a super-resolution algorithm and DTI images processed with a motion-tracked slice-to-volume registration algorithm. The DTI from individual subjects were used to generate 16 templates, each specific to a week of GA; this was accomplished by means of a tensor-to-tensor diffeomorphic deformable registration method integrated with kernel regression in age. Deterministic tractography was performed to outline the forceps major, forceps minor, bilateral corticospinal tracts (CST), bilateral inferior fronto-occipital fasciculus (IFOF), bilateral inferior longitudinal fasciculus (ILF), and bilateral uncinate fasciculus (UF). The mean fractional anisotropy (FA) and mean diffusivity (MD) was recorded for all tracts. For a subset of tracts (forceps major, CST and IFOF) we manually divided the tractograms into anatomy conforming segments to evaluate within-tract changes. We found tract-specific, non-linear, age related changes in FA and MD. Early in gestation, these trends appear to be dominated by cytoarchitectonic changes in the transient white matter fetal zones while later in gestation, trends conforming to the progression of myelination were observed. We also observed significant (local) heterogeneity in within-tract developmental trajectories for the CST, IFOF, and forceps major.

CloudReg: Automatic Terabyte-Scale Cross-Modal Brain Volume Registration

Quantifying terabyte-scale multi-modal human and animal imaging data requires scalable analysis tools. We developed CloudReg, an open-source, automatic, terabyte-scale, cloud-based image analysis pipeline that pre-processes and registers cross-modal volumetric datasets with artifacts via spatially-varying polynomial intensity transform. CloudReg accurately registers the following datasets to their respective atlases: in vivo human and ex vivo macaque brain magnetic resonance imaging, ex vivo mouse brain micro-computed tomography, and cleared murine brain light-sheet microscopy.