<p>This fileset is associated with the Left Atrial Segmentation Challenge 2013 (LASC'13... more <p>This fileset is associated with the Left Atrial Segmentation Challenge 2013 (LASC'13). LASC'13 was part of the STACOM'13 workshop, held in conjunction with MICCAI'13. Seven international research groups, comprising 11 algorithms, participated in the challenge.</p> <p>For a detailed report, please refer to:</p> <p>Tobon-Gomez C, Geers AJ, Peters, J, Weese J, Pinto K, Karim R, Ammar M, Daoudi A, Margeta J, Sandoval Z, Stender B, Zheng Y, Zuluaga, MA, Betancur J, Ayache N, Chikh MA, Dillenseger J-L, Kelm BM, Mahmoudi S, Ourselin S, Schlaefer A, Schaeffter T, Razavi R, Rhode KS. Benchmark for Algorithms Segmenting the Left Atrium From 3D CT and MRI Datasets. IEEE Transactions on Medical Imaging, 34(7):1460–1473, 2015.</p> <p>The challenge is also featured on <a href="http://www.cardiacatlas.org/challenges/left-atrium-segmentation-challenge/" target="_blank">http://www.cardiacatlas.org/challenges/left-atrium-segmentation-challenge/</a></p><p></p><p>The data and code of the challenge have been made publicly available to serve as a benchmark for left atrial segmentation algorithms. Code is hosted on <a href="https://github.com/catactg/lasc" target="_blank">https://github.com/catactg/lasc</a></p> <p>Feel free to contact us with any questions.</p> <p> </p> <p>This fileset consists of 10 MRI datasets for training segmentation algorithms. Included are the image and GT segmentation.</p> <p>gt_binary.mhd + gt_binary.raw: Binary image representation of GT</p> <p>image.mhd + image.raw: Image for training</p
The aortic valve area (AVA) and the pressure drop (PD) across the aortic valve are important quan... more The aortic valve area (AVA) and the pressure drop (PD) across the aortic valve are important quantities for characterizing an aortic valve stenosis. Using the Bernoulli equation and mass conservation, a relation between both quantities can be derived. We developed a simulation pipeline to assess the accuracy of this relation for realistic patient anatomies and blood flow rates. The key element of the pipeline is a shape-constrained deformable model (SCDM) for the segmentation of the aortic valve, the ascending aorta and the left ventricle over the cardiac cycle in cardiac CT images. Efficient segmentation enabled application of the simulation pipeline to cardiac CT image sequences of 22 patients. Planimetric AVA and Bernoulli-based PD estimates were computed from the same segmentation results. The resulting PD estimates show a high correlation (R = 0.97), but Bernoulli-based PD results are on average 25% smaller than the CFD-based results. The results contribute to a better understanding and interpretation of clinically used quantities such as the AVA and the PD.
This paper addresses the localization of anatomical structures in medical images by a Generalized... more This paper addresses the localization of anatomical structures in medical images by a Generalized Hough Transform (GHT). As localization is often a prerequisite for subsequent model-based segmentation, it is important to assess whether or not the GHT was able to locate the desired object. The GHT by its construction does not make this distinction. We present an approach to detect incorrect GHT localizations by deriving collective features of contributing GHT model points and by training a Support Vector Machine (SVM) classifier. On a training set of 204 cases, we demonstrate that for the detection of incorrect localizations classification errors of down to 3% are achievable. This is three times less than the observed intrinsic GHT localization error.
Dynamically modeling the word distribution in a variety of texts is a goal with various applicati... more Dynamically modeling the word distribution in a variety of texts is a goal with various applications. For speech recognition a dynamic unigram may efficiently be used for the adaptation of longer ranging language models. For information retrieval it may be a good starting point to predict the most characteristic words in document dependent queries. This short paper presents two approaches for adaptive unigram language models and illustrates their relation in a more general information theoretic framework.
One of the basic problems in modeling amorphous semiconductors is posed by the construction of a ... more One of the basic problems in modeling amorphous semiconductors is posed by the construction of a suitable geometry describing the positions of atoms and the network of bonds. Many attempts have been made to find reasonable models including hand-built ones and various computer-generated constructions. Not long ago a new approach was presented based upon an atomic decoration of the three-dimensional Penrose quasilattice. Therein, the structural elements form an icosahedral quasicrystalline tetracoordinated network. Its geometrical properties are analysed in detail in this paper. The electronic states of this quasicrystalline geometrical model for amorphous semiconductors are discussed in two following papers.
This is a repository copy of CDF-and Bernoulli-based pressure drop estimates: A comparison using ... more This is a repository copy of CDF-and Bernoulli-based pressure drop estimates: A comparison using patient anatomies from heart and aortic valve segmentation of CT images.
Proceedings of the 2003 Conference of the North American Chapter of the Association for Computational Linguistics on Human Language Technology companion volume of the Proceedings of HLT-NAACL 2003--short papers - NAACL '03, 2003
We investigate the optimal LM treatment of abundant filled pauses (FP) in spontaneous monologues ... more We investigate the optimal LM treatment of abundant filled pauses (FP) in spontaneous monologues of a professional dictation task. Questions addressed here are (1) how to deal with FP in the LM history and (2) to which extent can the LM distinguish between positions with high and low FP likelihood. Our results differ partly from observations reported on dialogues. Discarding FP from all LM histories clearly improves the performance. Local perplexities, entropies and word rankings at positions following FP suggest that most FP indicate hesitations rather than restarts. Proper prediction of FP allows to distinguish FP from word positions by a doubled FP probability. Recognition experiments confirm the improvements found in our perplexity studies.
An automatic procedure for detecting and segmenting anatomical objects in 3-D images is necessary... more An automatic procedure for detecting and segmenting anatomical objects in 3-D images is necessary for achieving a high level of automation in many medical applications. Since today's segmentation techniques typically rely on user input for initialization, they do not allow for a fully automatic workflow. In this work, the generalized Hough transform is used for detecting anatomical objects with well defined shape in 3-D medical images. This well-known technique has frequently been used for object detection in 2-D images and is known to be robust and reliable. However, its computational and memory requirements are generally huge, especially in case of considering 3-D images and various free transformation parameters. Our approach limits the complexity of the generalized Hough transform to a reasonable amount by (1) using object prior knowledge during the preprocessing in order to suppress unlikely regions in the image, (2) restricting the flexibility of the applied transformation to only scaling and translation, and (3) using a simple shape model which does not cover any inter-individual shape variability. Despite these limitations, the approach is demonstrated to allow for a coarse 3-D delineation of the femur, vertebra and heart in a number of experiments. Additionally it is shown that the quality of the object localization is in nearly all cases sufficient to initialize a successful segmentation using shape constrained deformable models.
<p>This fileset is associated with the Left Atrial Segmentation Challenge 2013 (LASC'13... more <p>This fileset is associated with the Left Atrial Segmentation Challenge 2013 (LASC'13). LASC'13 was part of the STACOM'13 workshop, held in conjunction with MICCAI'13. Seven international research groups, comprising 11 algorithms, participated in the challenge.</p> <p>For a detailed report, please refer to:</p> <p>Tobon-Gomez C, Geers AJ, Peters, J, Weese J, Pinto K, Karim R, Ammar M, Daoudi A, Margeta J, Sandoval Z, Stender B, Zheng Y, Zuluaga, MA, Betancur J, Ayache N, Chikh MA, Dillenseger J-L, Kelm BM, Mahmoudi S, Ourselin S, Schlaefer A, Schaeffter T, Razavi R, Rhode KS. Benchmark for Algorithms Segmenting the Left Atrium From 3D CT and MRI Datasets. IEEE Transactions on Medical Imaging, 34(7):1460–1473, 2015.</p> <p>The challenge is also featured on <a href="http://www.cardiacatlas.org/challenges/left-atrium-segmentation-challenge/" target="_blank">http://www.cardiacatlas.org/challenges/left-atrium-segmentation-challenge/</a></p><p></p><p>The data and code of the challenge have been made publicly available to serve as a benchmark for left atrial segmentation algorithms. Code is hosted on <a href="https://github.com/catactg/lasc" target="_blank">https://github.com/catactg/lasc</a></p> <p>Feel free to contact us with any questions.</p> <p> </p> <p>This fileset consists of 10 MRI datasets for training segmentation algorithms. Included are the image and GT segmentation.</p> <p>gt_binary.mhd + gt_binary.raw: Binary image representation of GT</p> <p>image.mhd + image.raw: Image for training</p
The aortic valve area (AVA) and the pressure drop (PD) across the aortic valve are important quan... more The aortic valve area (AVA) and the pressure drop (PD) across the aortic valve are important quantities for characterizing an aortic valve stenosis. Using the Bernoulli equation and mass conservation, a relation between both quantities can be derived. We developed a simulation pipeline to assess the accuracy of this relation for realistic patient anatomies and blood flow rates. The key element of the pipeline is a shape-constrained deformable model (SCDM) for the segmentation of the aortic valve, the ascending aorta and the left ventricle over the cardiac cycle in cardiac CT images. Efficient segmentation enabled application of the simulation pipeline to cardiac CT image sequences of 22 patients. Planimetric AVA and Bernoulli-based PD estimates were computed from the same segmentation results. The resulting PD estimates show a high correlation (R = 0.97), but Bernoulli-based PD results are on average 25% smaller than the CFD-based results. The results contribute to a better understanding and interpretation of clinically used quantities such as the AVA and the PD.
This paper addresses the localization of anatomical structures in medical images by a Generalized... more This paper addresses the localization of anatomical structures in medical images by a Generalized Hough Transform (GHT). As localization is often a prerequisite for subsequent model-based segmentation, it is important to assess whether or not the GHT was able to locate the desired object. The GHT by its construction does not make this distinction. We present an approach to detect incorrect GHT localizations by deriving collective features of contributing GHT model points and by training a Support Vector Machine (SVM) classifier. On a training set of 204 cases, we demonstrate that for the detection of incorrect localizations classification errors of down to 3% are achievable. This is three times less than the observed intrinsic GHT localization error.
Dynamically modeling the word distribution in a variety of texts is a goal with various applicati... more Dynamically modeling the word distribution in a variety of texts is a goal with various applications. For speech recognition a dynamic unigram may efficiently be used for the adaptation of longer ranging language models. For information retrieval it may be a good starting point to predict the most characteristic words in document dependent queries. This short paper presents two approaches for adaptive unigram language models and illustrates their relation in a more general information theoretic framework.
One of the basic problems in modeling amorphous semiconductors is posed by the construction of a ... more One of the basic problems in modeling amorphous semiconductors is posed by the construction of a suitable geometry describing the positions of atoms and the network of bonds. Many attempts have been made to find reasonable models including hand-built ones and various computer-generated constructions. Not long ago a new approach was presented based upon an atomic decoration of the three-dimensional Penrose quasilattice. Therein, the structural elements form an icosahedral quasicrystalline tetracoordinated network. Its geometrical properties are analysed in detail in this paper. The electronic states of this quasicrystalline geometrical model for amorphous semiconductors are discussed in two following papers.
This is a repository copy of CDF-and Bernoulli-based pressure drop estimates: A comparison using ... more This is a repository copy of CDF-and Bernoulli-based pressure drop estimates: A comparison using patient anatomies from heart and aortic valve segmentation of CT images.
Proceedings of the 2003 Conference of the North American Chapter of the Association for Computational Linguistics on Human Language Technology companion volume of the Proceedings of HLT-NAACL 2003--short papers - NAACL '03, 2003
We investigate the optimal LM treatment of abundant filled pauses (FP) in spontaneous monologues ... more We investigate the optimal LM treatment of abundant filled pauses (FP) in spontaneous monologues of a professional dictation task. Questions addressed here are (1) how to deal with FP in the LM history and (2) to which extent can the LM distinguish between positions with high and low FP likelihood. Our results differ partly from observations reported on dialogues. Discarding FP from all LM histories clearly improves the performance. Local perplexities, entropies and word rankings at positions following FP suggest that most FP indicate hesitations rather than restarts. Proper prediction of FP allows to distinguish FP from word positions by a doubled FP probability. Recognition experiments confirm the improvements found in our perplexity studies.
An automatic procedure for detecting and segmenting anatomical objects in 3-D images is necessary... more An automatic procedure for detecting and segmenting anatomical objects in 3-D images is necessary for achieving a high level of automation in many medical applications. Since today's segmentation techniques typically rely on user input for initialization, they do not allow for a fully automatic workflow. In this work, the generalized Hough transform is used for detecting anatomical objects with well defined shape in 3-D medical images. This well-known technique has frequently been used for object detection in 2-D images and is known to be robust and reliable. However, its computational and memory requirements are generally huge, especially in case of considering 3-D images and various free transformation parameters. Our approach limits the complexity of the generalized Hough transform to a reasonable amount by (1) using object prior knowledge during the preprocessing in order to suppress unlikely regions in the image, (2) restricting the flexibility of the applied transformation to only scaling and translation, and (3) using a simple shape model which does not cover any inter-individual shape variability. Despite these limitations, the approach is demonstrated to allow for a coarse 3-D delineation of the femur, vertebra and heart in a number of experiments. Additionally it is shown that the quality of the object localization is in nearly all cases sufficient to initialize a successful segmentation using shape constrained deformable models.
Uploads
Papers by Jochen Peters