Fast fully automatic brain detection in fetal MRI using dense rotation invariant image descriptors

Kainz, Bernhard; Keraudren, Kevin; Kyriakopoulou, Vanessa; Rutherford, Mary A.; Hajnal, Joseph V.; Rueckert, Daniel

doi:10.1109/isbi.2014.6868098

Cited by 25 publications

(22 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Two major types of approaches can be distinguished, either template-based segmentation [13][14][15] or machine learning [16][17][18] techniques. The first attempt 13 of fetal brain extraction proposed first to estimate the location of the eyes (based on rigid template registration) in order to segment the fetal brain using contrast, morphological and biometrical prior information.…”

Section: Description Of the Purposementioning

confidence: 99%

Automatic brain extraction in fetal MRI using multi-atlas-based segmentation

et al. 2015

View full text Add to dashboard Cite

In fetal brain MRI, most of the high-resolution reconstruction algorithms rely on brain segmentation as a preprocessing step. Manual brain segmentation is however highly time-consuming and therefore not a realistic solution. In this work, we assess on a large dataset the performance of Multiple Atlas Fusion (MAF) strategies to automatically address this problem. Firstly, we show that MAF significantly increase the accuracy of brain segmentation as regards single-atlas strategy. Secondly, we show that MAF compares favorably with the most recent approach (Dice above 0.90). Finally, we show that MAF could in turn provide an enhancement in terms of reconstruction quality. DESCRIPTION OF THE PURPOSEMost of the high-resolution reconstruction algorithms used in fetal MRI 1-10 rely only on brain tissue-relevant voxels of low-resolution (LR) images. In general, those algorithms need to perform brain segmentation as a preprocessing step. This brain extraction is essential to ensure good results of the subsequent image processing steps (denoising, bias correction, motion estimation, and super-resolution reconstruction). Despite of manual brain segmentation can be performed, it is highly time-consuming (around 15 minutes per stack of 15 slices) and not a realistic solution for large-scale studies. In the literature, accurate brain extraction tools have been developed for adult and infant brain MRI. 11, 12 But, fetal brain MRI differs a lot from neonatal or adult imaging in terms of image content (with maternal tissues surrounding the fetal brain), image contrast and brain size. Consequently, those tools are not well adapted to fetal MRI.Few works have addressed the automatic extraction of fetal brain in MRI. Two major types of approaches can be distinguished, either template-based segmentation 13-15 or machine learning 16-18 techniques. The first attempt 13 of fetal brain extraction proposed first to estimate the location of the eyes (based on rigid template registration) in order to segment the fetal brain using contrast, morphological and biometrical prior information. This method gave precise results in 22 out of 24 stacks of fetuses aged between 30 and 53 gestational weeks (GW). However, they relied on the assumption of low motion between slices that limits the robustness of the method to clinical databases where large motion can occur. More recently, Taleb et al. 14 presented an efficient brain extraction method based on single age-specific template segmentation (affine template registration) and fusion of orthogonal segmentations (transversal, coronal and sagittal) where final brain masks were successfully estimated in 82% of the cases. The validation was however rather qualitative (i.e. only success or failure label was given to the results). A supervised approach, 16 based on a two-phase random forest classifier, was adopted in order to obtain a method applicable to all fetal ages and more robust with respect to motion between slices. This method has shown comparable results to the method 13 but the whole brain was c...

show abstract

Section: Description Of the Purposementioning

confidence: 99%

Automatic brain extraction in fetal MRI using multi-atlas-based segmentation

et al. 2015

View full text Add to dashboard Cite

show abstract

“…This method first defines a region of interest (ROI) using the intersection of all scans and then registers this ROI to an age dependent fetal brain template, before refining the segmentation with a fusion step. Ison et al (2012) and Kainz et al (2014) address the variability of fetal MRI through the use of Machine Learning. Both methods rely on 3D features, either 3D Haar-like features or dense rotation invariant features, which are justified in cases of little motion, especially considering that the maternal anatomy is not moving and occupies most of the image.…”

Section: Brain Localization In Fetal Mrimentioning

confidence: 99%

Automated fetal brain segmentation from 2D MRI slices for motion correction

et al. 2014

View full text Add to dashboard Cite

Motion correction is a key element for imaging the fetal brain in-utero using Magnetic Resonance Imaging (MRI). Maternal breathing can introduce motion, but a larger effect is frequently due to fetal movement within the womb. Consequently, imaging is frequently performed slice-by-slice using single shot techniques, which are then combined into volumetric images using slice-tovolume reconstruction methods (SVR). For successful SVR, a key preprocessing step is to isolate fetal brain tissues from maternal anatomy before correcting for the motion of the fetal head. This has hitherto been a manual or semi-automatic procedure. We propose an automatic method to localize and segment the brain of the fetus when the image data is acquired as stacks of 2D slices with anatomy misaligned due to fetal motion. We combine this segmentation process with a robust motion correction method, enabling the segmentation to be refined as the reconstruction proceeds. The fetal brain localization process uses Maximally Stable Extremal Regions (MSER), which are classified using a Bag-of-Words model with Scale-Invariant Feature Transform (SIFT) features. The segmentation process is a patch-based propagation of the MSER regions selected during detection, combined with a Conditional Random Field (CRF). The gestational age (GA) is used to incorporate prior knowledge about the size and volume of the fetal brain into the detection and segmentation process. The method was tested in a ten-fold cross-validation experiment on 66 datasets of healthy fetuses whose GA ranged from 22 to 39 weeks. In 85% of the tested cases, our proposed method produced a motion corrected volume of a relevant quality for clinical diagnosis, thus removing the need for manually delineating the contours of the brain before motion correction. Our method automatically generated as a side-product a segmentation of the reconstructed fetal brain with a mean Dice score of 93%, which can be used for further processing.

show abstract

“…Often six to twelve stacks need to be acquired to sufficiently oversample a 3D volume. Segmentation and localization of selected organs can be automated, however, the available approaches provide either a very rough segmentation of the central slices of a stack [7] or require less motion corrupted stacks [4]. Furthermore, they are only applicable to a specificity trained region, e.g., the fetal brain.…”

Section: Introductionmentioning

confidence: 99%

Flexible Reconstruction and Correction of Unpredictable Motion from Stacks of 2D Images

Kainz

Alansary

Malamateniou

et al. 2015

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

Abstract. We present a method to correct motion in fetal in-utero scan sequences. The proposed approach avoids previously necessary manual segmentation of a region of interest. We solve the problem of non-rigid motion by splitting motion corrupted slices into overlapping patches of finite size. In these patches the assumption of rigid motion approximately holds and they can thus be used to perform a slice-to-volumebased (SVR) reconstruction during which their consistency with the other patches is learned. The learned information is used to reject patches that are not conform with the motion corrected reconstruction in their local areas. We evaluate rectangular and evenly distributed patches for the reconstruction as well as patches that have been derived from superpixels. Both approaches achieve on 29 subjects aged between 22-37 weeks a sufficient reconstruction quality and facilitate following 3D segmentation of fetal organs and the placenta.

show abstract

Fast fully automatic brain detection in fetal MRI using dense rotation invariant image descriptors

Cited by 25 publications

References 10 publications

Automatic brain extraction in fetal MRI using multi-atlas-based segmentation

Automatic brain extraction in fetal MRI using multi-atlas-based segmentation

Automated fetal brain segmentation from 2D MRI slices for motion correction

Flexible Reconstruction and Correction of Unpredictable Motion from Stacks of 2D Images

Contact Info

Product

Resources

About