A new method for robust organ positioning in CT images

Vik, T.; Bystrov, Daniel; Schadewaldt, Nicole; Schulz, Heinrich; Peters, Jörg

doi:10.1109/isbi.2012.6235553

Cited by 4 publications

(5 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For example, in a pelvic image, only few LUNG sample points will be chosen, whereas for a thorax image, many LUNG sample points are essential to provide a good match. Hence, the method proposed here overcomes the limitation of Vik et al 7 , that the image field-of-view has to be similar in the set of training images.…”

Section: Methodsmentioning

confidence: 96%

“…where the samples x i are drawn according to p(x|t), for details see Vik et al 7 It is essential, that the sampling is based on the image pdf, where probabilities are mostly 0 or 1, and is then compared to the transformed atlas for explanation. The singularity of the logarithmic function penalizes any image sample point, that cannot be explained by the atlas.…”

Section: Methodsmentioning

confidence: 99%

“…An extension to the method of Vik et al 7 is the dynamic determination of the number of sample points n t per tissue type t in (2), based on the ratio between feature responses in the atlas and in the image. In this way, the weighting of the tissue type influence adapts to the imaged anatomical region.…”

Section: Methodsmentioning

confidence: 99%

“…While Vik et al 7 construct the probabilistic tissue type atlas from registered training images based on a number of manually chosen ground truth landmarks, here a bootstrapping approach combines the registration step with the atlas creation: A single training image is used to construct a first tissue type atlas. Then all the training images are registered onto the initial atlas as described above.…”

Section: Methodsmentioning

confidence: 99%

“…1 Here we present two major contributions: a novel registration approach and an automatic atlas formation method. Following Vik et al 7 , the main idea there is to use a simple, efficient and effective tissue type classification which reduces the complexity of the registration problem and leads to fast and robust computations. The main contribution of this paper is a new strategy to dynamically determine the sample size related to tissue type estimations.…”

Section: Description Of Purposementioning

confidence: 99%

See 4 more Smart Citations

Annotation-free probabilistic atlas learning for robust anatomy detection in CT images

et al. 2015

Self Cite

View full text Add to dashboard Cite

A fully automatic method generating a whole body atlas from CT images is presented. The atlas serves as a reference space for annotations. It is based on a large collection of partially overlapping medical images and a registration scheme. The atlas itself consists of probabilistic tissue type maps and can represent anatomical variations. The registration scheme is based on an entropy-like measure of these maps and is robust with respect to field-of-view variations. In contrast to other atlas generation methods, which typically rely on a sufficiently large set of annotations on training cases, the presented method requires only the images. An iterative refinemen t strategy is used to automatically stitch the images to build the atlas.Affine registration of unseen CT images to the probabilistic atlas can be used to transfer reference annotations, e.g. organ models for segmentation initialization or reference bounding boxes for field-of-view selection. The robustness and generality of the method is shown using a three-fold cross-validation of the registration on a set of 316 CT images of unknown content and large anatomical variability. As an example, 17 organs are annotated in the atlas reference space and their localization in the test images is evaluated. The method yields a recall (sensitivity), specificity and precision of at least 96% and thus performs excellent in comparison to competitors. DESCRIPTION OF PURPOSEThe correlation of an unknown image to human anatomy is an important task in many medical image processing applications, 1-3 e.g. for automatic organ segmentation, which has applications in radiotherapy planning, surgery planning and other clinical tasks. Other applications are workflow simplification by detection of image content and automation of subsequent processing steps or large scale data analysis by retrieving certain anatomical regions from image databases.Two approaches to anatomy correlation can be distinguished: structure detection and registration to a reference space. A recent detection approach for organs has been published by Criminisi et al. 4 , where trained regression forests are used for bounding box detection. The application to an unseen image is fast, robust, and organ localization accuracy outperforms standard registration-based methods. A prerequisite is the definition of tight organ bounding boxes in all training data sets. The detection result is then also a bounding box.While the detection of individual structures must typically be re-trained for new organs, registration to a reference space can be used to detect the field-of-view and estimate the locations of several organs simultaneously. Multi-atlas approaches 5 are able to handle a variety of images via a database of reference images which are all registered to an unknown image independently. These approaches suffer from long computation times and a strong dependency on the selection of example images in the atlas. In contrast, atlases with a single reference image have difficulties to deal with local variability.In r...

show abstract

Section: Methodsmentioning

confidence: 96%

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Description Of Purposementioning

confidence: 99%

See 3 more Smart Citations

Annotation-free probabilistic atlas learning for robust anatomy detection in CT images

et al. 2015

Self Cite

View full text Add to dashboard Cite

show abstract

State‐of‐the‐Art Report: Visual Computing in Radiation Therapy Planning

Schlachter

Raidou

Muren

et al. 2019

Computer Graphics Forum

View full text Add to dashboard Cite

Radiation therapy (RT) is one of the major curative approaches for cancer. It is a complex and risky treatment approach, which requires precise planning, prior to the administration of the treatment. Visual Computing (VC) is a fundamental component of RT planning, providing solutions in all parts of the process—from imaging to delivery. Despite the significant technological advancements of RT over the last decades, there are still many challenges to address. This survey provides an overview of the compound planning process of RT, and of the ways that VC has supported RT in all its facets. The RT planning process is described to enable a basic understanding in the involved data, users and workflow steps. A systematic categorization and an extensive analysis of existing literature in the joint VC/RT research is presented, covering the entire planning process. The survey concludes with a discussion on lessons learnt, current status, open challenges, and future directions in VC/RT research.

show abstract