Automated ventricular systems segmentation in brain CT images by combining low-level segmentation and high-level template matching

Smith, Rebecca; Ji, Soo Yeon; Ward, Kevin R.; Najarian, Kayvan

doi:10.1186/1472-6947-9-s1-s4

Cited by 53 publications

(51 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Although dozens of CT images can be acquired from a patient's brain scan, only those slices depicting clear anatomical features and limited inherent noise are used in the IML quantification. Currently in clinical practice, the process of selecting appropriate images for IML diagnosis is performed manually by physicians [7,8]. From our research, we did not find any existing automated method that is in implementation to perform this task.…”

Section: Introductionmentioning

confidence: 70%

“…More details can be found in [8]. Finally, the rotation angle θ with the minimum symmetry cost S θ determines the rotation direction of the midline of the brain for each particular CT slice.…”

Section: Approximate Iml Detectionmentioning

confidence: 99%

“…In the calculation of the SSA algorithm, the detected skull  and its mass center (x , y ) have been determined y Equ b proximate ideal midline, we use the e ric position search algorithm which was developed for a prior work by our research group [8].…”

Section: Approximate Iml Detectionmentioning

confidence: 99%

“…G. Ruppert et al extracted the MSP based on bilateral symmetry maximization [6]. W. Chen et al combined bone symmetry and direct detection of the anatomical features in CT images in IML detection [8]. This method works effectively and accurately on a single CT slice but lacks connection or comparison with the detection results from other CT slices of the same patient.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Ideal Midline Detection Using Automated Processing of Brain CT Image

Qi¹,

Belle²,

Shandilya³

et al. 2013

OJMI

Self Cite

View full text Add to dashboard Cite

Brain ideal midline estimation is vital in medical image processing, especially in analyzing the severity of a brain injury in clinical environments. We propose an automated computer-aided ideal midline estimation system with a two-step process. First, a CT Slice Selection Algorithm (SSA) can automatically select an appropriate subset of slices from a large number of raw CT images using the skull's anatomical features. Next, an ideal midline detection is implemented on the selected subset of slices. An exhaustive symmetric position search is performed based on the anatomical features in the detection. In order to enhance the accuracy of the detection, a global rotation assumption is applied to determine the ideal midline by fully considering the connection between slices. Experimental results of the multi-stage algorithm were assessed on 3313 CT slices of 70 patients. The accuracy of the proposed system is 96.9%, which makes it viable for use under clinical settings.

show abstract

Section: Introductionmentioning

confidence: 70%

Section: Approximate Iml Detectionmentioning

confidence: 99%

Section: Approximate Iml Detectionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Ideal Midline Detection Using Automated Processing of Brain CT Image

Qi¹,

Belle²,

Shandilya³

et al. 2013

OJMI

Self Cite

View full text Add to dashboard Cite

show abstract

“…First, the ideal midline of the brain, i.e. the midline before injury, is found via a hierarchical search based on skull symmetry and tissue features 3 . Secondly, the ventricular system is segmented for each brain CT image 4 .…”

Section: Methodology Overviewmentioning

confidence: 99%

Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images

Belle

Cockrell

Ward

et al. 2013

JoVE

Self Cite

View full text Add to dashboard Cite

In this paper we present an automated system based mainly on the computed tomography (CT) images consisting of two main components: the midline shift estimation and intracranial pressure (ICP) pre-screening system. To estimate the midline shift, first an estimation of the ideal midline is performed based on the symmetry of the skull and anatomical features in the brain CT scan. Then, segmentation of the ventricles from the CT scan is performed and used as a guide for the identification of the actual midline through shape matching. These processes mimic the measuring process by physicians and have shown promising results in the evaluation. In the second component, more features are extracted related to ICP, such as the texture information, blood amount from CT scans and other recorded features, such as age, injury severity score to estimate the ICP are also incorporated. Machine learning techniques including feature selection and classification, such as Support Vector Machines (SVMs), are employed to build the prediction model using RapidMiner. The evaluation of the prediction shows potential usefulness of the model. The estimated ideal midline shift and predicted ICP levels may be used as a fast pre-screening step for physicians to make decisions, so as to recommend for or against invasive ICP monitoring. Video LinkThe video component of this article can be found at

show abstract

Bone segmentation and 3D visualization of CT images for traumatic pelvic injuries

Belle

Hargraves

et al. 2014

Int J Imaging Syst Tech

Self Cite

View full text Add to dashboard Cite

Pelvic bone segmentation is a vital step in analyzing pelvic CT images, which assists physicians with diagnostic decision making in cases of traumatic pelvic injuries. Due to the limited resolution of the original CT images and the complexity of pelvic structures and their possible fractures, automatic pelvic bone segmentation in multiple CT slices is very difficult. In this study, an automatic pelvic bone segmentation approach is proposed using the combination of anatomical knowledge and computational techniques. It is developed for solving the problem of accurate and efficient bone segmentation using multiple consecutive pelvic CT slices obtained from each patient. Our proposed segmentation method is able to handle variation of bone shapes between slices there by making it less susceptible to inter‐personal variability between different patients' data. Moreover, the designed training models are validated using a cross‐validation process to demonstrate the effectiveness. The algorithm's capability is tested on a set of 20 CT data sets. Successful segmentation results and quantitative evaluations are present to demonstrate the effectiveness and robustness of proposed algorithm, well suited for pelvic bone segmentation purposes.

show abstract

Automated ventricular systems segmentation in brain CT images by combining low-level segmentation and high-level template matching

Cited by 53 publications

References 24 publications

Ideal Midline Detection Using Automated Processing of Brain CT Image

Ideal Midline Detection Using Automated Processing of Brain CT Image

Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images

Bone segmentation and 3D visualization of CT images for traumatic pelvic injuries

Contact Info

Product

Resources

About