Whole Brain Parcellation with Pathology: Validation on Ventriculomegaly Patients

Carass, Aaron; Shao, Muhan; Li, Xiang; Blitz, Ari M.; Roy, Snehashis; Pham, Dzung L.; Prince, Jerry L.; Ellingsen, Lotta M.

doi:10.1007/978-3-319-67434-6_3

Cited by 10 publications

(9 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Automated parcellation has been shown to remove manual bias, thus increasing consistency and comparability across studies and sites (Buckner et al, 2004; Hsu et al, 2002). We are only concerned with those methods that provide at a minimum the lobes of the cerebellum; hence methods like FreeSurfer (Dale et al, 1999; Fischl et al, 2004; Reuter and Fischl, 2011; Fischl, 2012), BrainSuite (Shattuck et al, 2001; Shattuck and Leahy, 2002; Shattuck et al, 2008), TOADS (Bazin and Pham, 2008; Shiee et al, 2010), CRUISE (Tosun et al, 2006; Landman et al, 2013), CRUISE+ (Shiee et al, 2014), MA-CRUISE (Huo et al, 2016a,b), and others (Carass et al, 2017c; Desikan et al, 2006; Guo et al, 2017; Ledig et al, 2015; Liu et al, 2012; Roy et al, 2015; Shao et al, 2018; Shiee et al, 2011; Tomas-Fernandez and Warfield, 2015; Van Leemput et al, 1999; Zhang et al, 2001; Zhao et al, 2017) that only provide tissue classes or coarse parcellations of the cerebellum are not directly relevant unless used in combination with other tools. The first published method that provided a fully automated parcellation of the cerebellar lobules was SUIT (Diedrichsen, 2006); the method used a spatially unbiased template of the human cerebellum that when registered with a subject image provided the parcellation.…”

Section: Introductionmentioning

confidence: 99%

Comparing fully automated state-of-the-art cerebellum parcellation from magnetic resonance images

Carass

Cuzzocreo²,

Han

et al. 2018

NeuroImage

Self Cite

View full text Add to dashboard Cite

The human cerebellum plays an essential role in motor control, is involved in cognitive function (i.e., attention, working memory, and language), and helps to regulate emotional responses. Quantitative in-vivo assessment of the cerebellum is important in the study of several neurological diseases including cerebellar ataxia, autism, and schizophrenia. Different structural subdivisions of the cerebellum have been shown to correlate with differing pathologies. To further understand these pathologies, it is helpful to automatically parcellate the cerebellum at the highest fidelity possible. In this paper, we coordinated with colleagues around the world to evaluate automated cerebellum parcellation algorithms on two clinical cohorts showing that the cerebellum can be parcellated to a high accuracy by newer methods. We characterize these various methods at four hierarchical levels: coarse (i.e., whole cerebellum and gross structures), lobe, subdivisions of the vermis, and the lobules. Due to the number of labels, the hierarchy of labels, the number of algorithms, and the two cohorts, we have restricted our analyses to the Dice measure of overlap. Under these conditions, machine learning based methods provide a collection of strategies that are efficient and deliver parcellations of a high standard across both cohorts, surpassing previous work in the area. In conjunction with the rank-sum computation, we identified an overall winning method.

show abstract

Section: Introductionmentioning

confidence: 99%

Comparing fully automated state-of-the-art cerebellum parcellation from magnetic resonance images

Carass

Cuzzocreo²,

Han

et al. 2018

NeuroImage

Self Cite

View full text Add to dashboard Cite

show abstract

“…The proposed method is compared to a multi-atlas segmentation method, RUDOLPH (RobUst DictiOnary- learning and Label Propagation Hybrid), 11,12 which is designed for subjects with enlarged ventricles and labels cortical CSF on a voxel grid. The Dice coefficient between the semi-automatic delineation and automatic methods were computed, which is given by Dice=2|A∩B||A|+|B|.…”

Section: Experiments and Resultsmentioning

confidence: 99%

Deformable model reconstruction of the subarachnoid space

Glaister

Shao

et al. 2018

Medical Imaging 2018: Image Processing

Self Cite

View full text Add to dashboard Cite

The subarachnoid space is a layer in the meninges that surrounds the brain and is filled with trabeculae and cerebrospinal fluid. Quantifying the volume and thickness of the subarachnoid space is of interest in order to study the pathogenesis of neurodegenerative diseases and compare with healthy subjects. We present an automatic method to reconstruct the subarachnoid space with subvoxel accuracy using a nested deformable model. The method initializes the deformable model using the convex hull of the union of the outer surfaces of the cerebrum, cerebellum and brainstem. A region force is derived from the subject’s Tl-weighted and T2-weighted MRI to drive the deformable model to the outer surface of the subarachnoid space. The proposed method is compared to a semi-automatic delineation from the subject’s T2-weighted MRI and an existing multi-atlas-based method. A small pilot study comparing the volume and thickness measurements in a set of age-matched subjects with normal pressure hydrocephalus and healthy controls is presented to show the efficacy of the proposed method.

show abstract

“…The performance of the VParNet was compared with four state-of-the-art brain segmentation methods: FreeSurfer (version 6.0.0) ( Dale et al, 1999 ; Fischl et al, 2002 ; Fischl, 2012 ), MALPEM ( Ledig et al, 2015 ), Joint label fusion (JLF) from the ANTs software package ( Wang et al, 2013 ; Wang and Yushkevich, 2013 ), and RUDOLPH ( Ellingsen et al, 2016 ; Carass et al, 2017 ; Shao et al, 2018a ). We also conducted an ablation analysis of the VParNet to see how different strategies affect the performance.…”

Section: Methodsmentioning

confidence: 99%

“…One reason for some of these failures is that most methods rely on registration between the atlas and subject images, which in pathological cases is rarely optimal. One recent algorithm, called RUDOLPH ( Ellingsen et al, 2016 ; Carass et al, 2017 ; Shao et al, 2018a ), was specifically designed to segment subjects with enlarged ventricles. Although robust in ventricle parcellation, this method takes hours to process a single image.…”

Section: Introductionmentioning

confidence: 99%

“… Joint label fusion ( Wang et al, 2013 ) ✓ ✓ Multi-atlas label fusion method formulating a weighted voting scheme to minimize the total expectation of the labeling error. RUDOLPH ( Ellingsen et al, 2016 ; Carass et al, 2017 ; Shao et al, 2018a ) ✓ ✓ Combines tissue segmentation and multi-atlas segmentation to correct registration priors. Designed for subjects with ventriculomegaly.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Brain ventricle parcellation using a deep neural network: Application to patients with ventriculomegaly

Shao

Han

Carass

et al. 2019

NeuroImage: Clinical

Self Cite

View full text Add to dashboard Cite

Numerous brain disorders are associated with ventriculomegaly, including both neuro-degenerative diseases and cerebrospinal fluid disorders. Detailed evaluation of the ventricular system is important for these conditions to help understand the pathogenesis of ventricular enlargement and elucidate novel patterns of ventriculomegaly that can be associated with different diseases. One such disease is normal pressure hydrocephalus (NPH), a chronic form of hydrocephalus in older adults that causes dementia. Automatic parcellation of the ventricular system into its sub-compartments in patients with ventriculomegaly is quite challenging due to the large variation of the ventricle shape and size. Conventional brain labeling methods are time-consuming and often fail to identify the boundaries of the enlarged ventricles. We propose a modified 3D U-Net method to perform accurate ventricular parcellation, even with grossly enlarged ventricles, from magnetic resonance images (MRIs). We validated our method on a data set of healthy controls as well as a cohort of 95 patients with NPH with mild to severe ventriculomegaly and compared with several state-of-the-art segmentation methods. On the healthy data set, the proposed network achieved mean Dice similarity coefficient (DSC) of 0.895 ± 0.03 for the ventricular system. On the NPH data set, we achieved mean DSC of 0.973 ± 0.02, which is significantly ( p < 0.005) higher than four state-of-the-art segmentation methods we compared with. Furthermore, the typical processing time on CPU-base implementation of the proposed method is 2 min, which is much lower than the several hours required by the other methods. Results indicate that our method provides: 1) highly robust parcellation of the ventricular system that is comparable in accuracy to state-of-the-art methods on healthy controls; 2) greater robustness and significantly more accurate results on cases of ventricular enlargement; and 3) a tool that enables computation of novel imaging biomarkers for dilated ventricular spaces that characterize the ventricular system.

show abstract

Whole Brain Parcellation with Pathology: Validation on Ventriculomegaly Patients

Cited by 10 publications

References 19 publications

Comparing fully automated state-of-the-art cerebellum parcellation from magnetic resonance images

Comparing fully automated state-of-the-art cerebellum parcellation from magnetic resonance images

Deformable model reconstruction of the subarachnoid space

Brain ventricle parcellation using a deep neural network: Application to patients with ventriculomegaly

Contact Info

Product

Resources

About