Pub Date : 2010-04-14DOI: 10.1109/ISBI.2010.5490264
A. Madabhushi, A. Basavanhally, Scott Doyle, S. Agner, George Lee
Computer-aided prognosis (CAP) is a new and exciting complement to the field of computer-aided diagnosis (CAD) and involves developing computerized image analysis and multi-modal data fusion algorithms for helping physicians predict disease outcome and patient survival. At the Laboratory for Computational Imaging and Bioinformatics (LCIB)1 at Rutgers University we have been developing computerized algorithms for high dimensional data and image analysis for predicting disease outcome from multiple modalities includng MRI, digital pathology, and protein expression. Additionally, we have been developing novel data fusion algorithms based on nonlinear dimensionality reduction methods (such as Graph Embedding) to quantitatively integrate prognostic information from multiple data sources and modalities. In this paper, we briefly describe 5 representative and ongoing CAP projects at LCIB. These projects include (1) an Image-based Risk Score (IbRiS) algorithm for predicting outcome of ER+ breast cancer patients based on quantitative image analysis of digitized breast cancer biopsy specimens alone, (2) segmenting and determining extent of lymphocytic infiltration (identified as a possible prognostic marker for outcome in Her2+ breast cancers) from digitized histopathology, (3) segmenting and diagnosing highly agressive triple-negative breast cancers on dynamic contrast enhanced (DCE) MRI, (4) distinguishing patients with different Gleason grades of prostate cancer (grade being known to be correlated to outcome) from digitzed needle biopsy specimens, and (5) integrating protein expression measurements obtained from mass spectrometry with quantitative image features derived from digitized histopathology for distinguishing between prostate cancer patients at low and high risk of disease recurrence.
{"title":"Computer-aided prognosis: Predicting patient and disease outcome via multi-modal image analysis","authors":"A. Madabhushi, A. Basavanhally, Scott Doyle, S. Agner, George Lee","doi":"10.1109/ISBI.2010.5490264","DOIUrl":"https://doi.org/10.1109/ISBI.2010.5490264","url":null,"abstract":"Computer-aided prognosis (CAP) is a new and exciting complement to the field of computer-aided diagnosis (CAD) and involves developing computerized image analysis and multi-modal data fusion algorithms for helping physicians predict disease outcome and patient survival. At the Laboratory for Computational Imaging and Bioinformatics (LCIB)1 at Rutgers University we have been developing computerized algorithms for high dimensional data and image analysis for predicting disease outcome from multiple modalities includng MRI, digital pathology, and protein expression. Additionally, we have been developing novel data fusion algorithms based on nonlinear dimensionality reduction methods (such as Graph Embedding) to quantitatively integrate prognostic information from multiple data sources and modalities. In this paper, we briefly describe 5 representative and ongoing CAP projects at LCIB. These projects include (1) an Image-based Risk Score (IbRiS) algorithm for predicting outcome of ER+ breast cancer patients based on quantitative image analysis of digitized breast cancer biopsy specimens alone, (2) segmenting and determining extent of lymphocytic infiltration (identified as a possible prognostic marker for outcome in Her2+ breast cancers) from digitized histopathology, (3) segmenting and diagnosing highly agressive triple-negative breast cancers on dynamic contrast enhanced (DCE) MRI, (4) distinguishing patients with different Gleason grades of prostate cancer (grade being known to be correlated to outcome) from digitzed needle biopsy specimens, and (5) integrating protein expression measurements obtained from mass spectrometry with quantitative image features derived from digitized histopathology for distinguishing between prostate cancer patients at low and high risk of disease recurrence.","PeriodicalId":250523,"journal":{"name":"2010 IEEE International Symposium on Biomedical Imaging: From Nano to Macro","volume":"112 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2010-04-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124841417","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2010-04-14DOI: 10.1109/ISBI.2010.5490113
J. Frikel
We propose a new framework for limited angle tomographic reconstruction. Our approach is based on the observation that for a given acquisition geometry only a few (visible) structures of the object can be reconstructed reliably using a limited angle data set. By formulating this problem in the curvelet domain, we can characterize those curvelet coefficients which correspond to visible structures in the image domain. The integration of this information into the formulation of the reconstruction problem leads to a considerable dimensionality reduction and yields a speedup of the corresponding reconstruction algorithms.
{"title":"A new framework for sparse regularization in limited angle x-ray tomography","authors":"J. Frikel","doi":"10.1109/ISBI.2010.5490113","DOIUrl":"https://doi.org/10.1109/ISBI.2010.5490113","url":null,"abstract":"We propose a new framework for limited angle tomographic reconstruction. Our approach is based on the observation that for a given acquisition geometry only a few (visible) structures of the object can be reconstructed reliably using a limited angle data set. By formulating this problem in the curvelet domain, we can characterize those curvelet coefficients which correspond to visible structures in the image domain. The integration of this information into the formulation of the reconstruction problem leads to a considerable dimensionality reduction and yields a speedup of the corresponding reconstruction algorithms.","PeriodicalId":250523,"journal":{"name":"2010 IEEE International Symposium on Biomedical Imaging: From Nano to Macro","volume":"44 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2010-04-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121719860","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2010-04-14DOI: 10.1109/ISBI.2010.5490080
B. M. Eyüboğlu, V. Arpinar, R. Boyacioglu, E. Değirmenci, G. Eker
Several algorithms have been proposed for image reconstruction in MREIT. These algorithms reconstruct conductivity distribution either directly from magnetic flux density measurements or from reconstructed current density distribution. In this study, performance of all major algorithms are evaluated and compared on a common platform, in terms of their reconstruction error, reconstruction time, perceptual image quality, immunity against measurement noise, required electrode size. J-Substitution (JS) and Hybrid J-Substitution algorithms have the best reconstruction accuracy but they are among the slowest. Another current density based algorithm, Equipotential Projection (EPP) algorithm along with magnetic flux density based Bz Sensitivity (BzS) algorithm has moderate reconstruction accuracy. BzS algorithm is the fastest.
{"title":"Comparison of magnetic resonance electrical impedance tomography (MREIT) reconstruction algorithms","authors":"B. M. Eyüboğlu, V. Arpinar, R. Boyacioglu, E. Değirmenci, G. Eker","doi":"10.1109/ISBI.2010.5490080","DOIUrl":"https://doi.org/10.1109/ISBI.2010.5490080","url":null,"abstract":"Several algorithms have been proposed for image reconstruction in MREIT. These algorithms reconstruct conductivity distribution either directly from magnetic flux density measurements or from reconstructed current density distribution. In this study, performance of all major algorithms are evaluated and compared on a common platform, in terms of their reconstruction error, reconstruction time, perceptual image quality, immunity against measurement noise, required electrode size. J-Substitution (JS) and Hybrid J-Substitution algorithms have the best reconstruction accuracy but they are among the slowest. Another current density based algorithm, Equipotential Projection (EPP) algorithm along with magnetic flux density based Bz Sensitivity (BzS) algorithm has moderate reconstruction accuracy. BzS algorithm is the fastest.","PeriodicalId":250523,"journal":{"name":"2010 IEEE International Symposium on Biomedical Imaging: From Nano to Macro","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2010-04-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114944389","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2010-04-14DOI: 10.1109/ISBI.2010.5490389
Hang Chang, B. Parvin
Membrane-bound macromolecules play an important role in tissue architecture and cell-cell communication, and is regulated by almost one-third of the genome. At the optical scale, one group of membrane proteins expresses themselves as linear structures along the cell surface boundaries, while others are sequestered. This paper targets the former group, whose intensity distributions are often heterogeneous and may lack specificity. Segmentation of the membrane protein enables the quantitative assessment of localization for comparative analysis. We introduce a three-step process to (i) regularize the membrane signal through iterative tangential voting, (ii) constrain the location of surface proteins by nuclear features, and (iii) assign membrane proteins to individual cells through an application of multi-phase geodesic level-set. We have validated our method against a dataset of 200 images, and demonstrated that multiphase level set has a superior performance compared to gradient vector flow snake.
{"title":"Multiphase level set for automated delineation of membrane-bound macromolecules","authors":"Hang Chang, B. Parvin","doi":"10.1109/ISBI.2010.5490389","DOIUrl":"https://doi.org/10.1109/ISBI.2010.5490389","url":null,"abstract":"Membrane-bound macromolecules play an important role in tissue architecture and cell-cell communication, and is regulated by almost one-third of the genome. At the optical scale, one group of membrane proteins expresses themselves as linear structures along the cell surface boundaries, while others are sequestered. This paper targets the former group, whose intensity distributions are often heterogeneous and may lack specificity. Segmentation of the membrane protein enables the quantitative assessment of localization for comparative analysis. We introduce a three-step process to (i) regularize the membrane signal through iterative tangential voting, (ii) constrain the location of surface proteins by nuclear features, and (iii) assign membrane proteins to individual cells through an application of multi-phase geodesic level-set. We have validated our method against a dataset of 200 images, and demonstrated that multiphase level set has a superior performance compared to gradient vector flow snake.","PeriodicalId":250523,"journal":{"name":"2010 IEEE International Symposium on Biomedical Imaging: From Nano to Macro","volume":"18 10","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2010-04-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114067055","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2010-04-14DOI: 10.1109/ISBI.2010.5490365
M. Guerquin-Kern, F. I. Karahanoğlu, D. Ville, K. Pruessmann, M. Unser
We present an analytical form of ground-truth k-space data for the 2-D Shepp-Logan brain phantom in the presence of multiple and non-homogeneous receiving coils. The analytical form allows us to conduct realistic simulations and validations of reconstruction algorithms for parallel MRI. The key contribution of our work is to use a polynomial representation of the coil's sensitivity. We show that this method is particularly accurate and fast with respect to the conventional methods. The implementation is made available to the community.
{"title":"Analytical form of Shepp-Logan phantom for parallel MRI","authors":"M. Guerquin-Kern, F. I. Karahanoğlu, D. Ville, K. Pruessmann, M. Unser","doi":"10.1109/ISBI.2010.5490365","DOIUrl":"https://doi.org/10.1109/ISBI.2010.5490365","url":null,"abstract":"We present an analytical form of ground-truth k-space data for the 2-D Shepp-Logan brain phantom in the presence of multiple and non-homogeneous receiving coils. The analytical form allows us to conduct realistic simulations and validations of reconstruction algorithms for parallel MRI. The key contribution of our work is to use a polynomial representation of the coil's sensitivity. We show that this method is particularly accurate and fast with respect to the conventional methods. The implementation is made available to the community.","PeriodicalId":250523,"journal":{"name":"2010 IEEE International Symposium on Biomedical Imaging: From Nano to Macro","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2010-04-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124173537","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2010-04-14DOI: 10.1109/ISBI.2010.5490163
O. Acosta, J. Fripp, A. Rueda, D. Xiao, E. Bonner, P. Bourgeat, Olivier Salvado
Deformable registration of cortical surfaces facilitates longitudinal and intergroup comparisons of cortical structure and function in the study of many neurodegenerative diseases. Non-rigid cortical matching is a challenging task due to the large variability between individuals and the complexity of the cortex. We present a new framework for computing cortical correspondences on brain surfaces based on 3D Shape Context and mean curvatures of partially flattened surfaces (PFS). Our approach is scale invariant and provides an accurate and anatomically meaningful alignment across the population. Registering PFS, instead of original cortical surfaces, simplifies the determination of shape correspondences, overcoming the problem of intersubject variability, while still guaranteeing the alignment of the main brain lobes and folding patterns. We validated the approach using 30 segmented brains from the OASIS database registered to a common space and compared the results with Freesurfer. In average, mean absolute distance of 0.36 and Hausdorff distance of 5.06 between moving and target surfaces are obtained. Further localization of labelled areas on each hemisphere demonstrated the accuracy of the technique.
{"title":"3D shape context surface registration for cortical mapping","authors":"O. Acosta, J. Fripp, A. Rueda, D. Xiao, E. Bonner, P. Bourgeat, Olivier Salvado","doi":"10.1109/ISBI.2010.5490163","DOIUrl":"https://doi.org/10.1109/ISBI.2010.5490163","url":null,"abstract":"Deformable registration of cortical surfaces facilitates longitudinal and intergroup comparisons of cortical structure and function in the study of many neurodegenerative diseases. Non-rigid cortical matching is a challenging task due to the large variability between individuals and the complexity of the cortex. We present a new framework for computing cortical correspondences on brain surfaces based on 3D Shape Context and mean curvatures of partially flattened surfaces (PFS). Our approach is scale invariant and provides an accurate and anatomically meaningful alignment across the population. Registering PFS, instead of original cortical surfaces, simplifies the determination of shape correspondences, overcoming the problem of intersubject variability, while still guaranteeing the alignment of the main brain lobes and folding patterns. We validated the approach using 30 segmented brains from the OASIS database registered to a common space and compared the results with Freesurfer. In average, mean absolute distance of 0.36 and Hausdorff distance of 5.06 between moving and target surfaces are obtained. Further localization of labelled areas on each hemisphere demonstrated the accuracy of the technique.","PeriodicalId":250523,"journal":{"name":"2010 IEEE International Symposium on Biomedical Imaging: From Nano to Macro","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2010-04-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126379028","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2010-04-14DOI: 10.1109/ISBI.2010.5490429
C. I. Sánchez, M. Niemeijer, M. Suttorp-Schulten, M. Abràmoff, B. Ginneken
Contextual information is of paramount importance in medical image understanding to detect and differentiate pathologies, especially when interpreting difficult cases. Current computer-aided detection (CAD) systems typically employ only local information to classify candidates, without taking into account global image information or the relation of a candidate with neighboring structures. In this work, we improve the detection of hard exudates in retinal images incorporating contextual information in the CAD system. The context is described by means of high-level contextual-based features based on the spatial relation with surrounding anatomical landmarks and similar lesions. Results show that a contextual CAD system for hard exudate detection is superior to an approach that uses only local information, with a significant increase of the figure of merit of the Free Receiver Operating Characteristic (FROC) curve from 0.840 to 0.945.
{"title":"Improving hard exudate detection in retinal images through a combination of local and contextual information","authors":"C. I. Sánchez, M. Niemeijer, M. Suttorp-Schulten, M. Abràmoff, B. Ginneken","doi":"10.1109/ISBI.2010.5490429","DOIUrl":"https://doi.org/10.1109/ISBI.2010.5490429","url":null,"abstract":"Contextual information is of paramount importance in medical image understanding to detect and differentiate pathologies, especially when interpreting difficult cases. Current computer-aided detection (CAD) systems typically employ only local information to classify candidates, without taking into account global image information or the relation of a candidate with neighboring structures. In this work, we improve the detection of hard exudates in retinal images incorporating contextual information in the CAD system. The context is described by means of high-level contextual-based features based on the spatial relation with surrounding anatomical landmarks and similar lesions. Results show that a contextual CAD system for hard exudate detection is superior to an approach that uses only local information, with a significant increase of the figure of merit of the Free Receiver Operating Characteristic (FROC) curve from 0.840 to 0.945.","PeriodicalId":250523,"journal":{"name":"2010 IEEE International Symposium on Biomedical Imaging: From Nano to Macro","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2010-04-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129646837","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2010-04-14DOI: 10.1109/ISBI.2010.5490138
N. Honnorat, Régis Vaillant, N. Paragios
Fluroscopic imaging provides means to assess the motion of the internal structures and therefore is of great use during surgery. In this paper we propose a novel approach for the segmentation of curvilinear structures in these images. The main challenge to be addressed is the lack of visual support due to the low SNR where traditional edge-based methods fail. Our approach combines machine learning techniques, unsupervised clustering and linear programming. In particular, numerous invariant to position/rotation classifiers are combined to detect candidate pixels of curvilinear structure. These candidates are grouped into consistent geometric segments through the use of a state-of-the art unsupervised clustering algorithm. The complete curvilinear structure is obtained through an ordering of these segments using the elastica model in a linear programming framework. Very promising results were obtained on guide wire segmentation in fluoroscopic images.
{"title":"Robust guidewire segmentation through boosting, clustering and linear programming","authors":"N. Honnorat, Régis Vaillant, N. Paragios","doi":"10.1109/ISBI.2010.5490138","DOIUrl":"https://doi.org/10.1109/ISBI.2010.5490138","url":null,"abstract":"Fluroscopic imaging provides means to assess the motion of the internal structures and therefore is of great use during surgery. In this paper we propose a novel approach for the segmentation of curvilinear structures in these images. The main challenge to be addressed is the lack of visual support due to the low SNR where traditional edge-based methods fail. Our approach combines machine learning techniques, unsupervised clustering and linear programming. In particular, numerous invariant to position/rotation classifiers are combined to detect candidate pixels of curvilinear structure. These candidates are grouped into consistent geometric segments through the use of a state-of-the art unsupervised clustering algorithm. The complete curvilinear structure is obtained through an ordering of these segments using the elastica model in a linear programming framework. Very promising results were obtained on guide wire segmentation in fluoroscopic images.","PeriodicalId":250523,"journal":{"name":"2010 IEEE International Symposium on Biomedical Imaging: From Nano to Macro","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2010-04-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129766996","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2010-04-14DOI: 10.1109/ISBI.2010.5490135
A. Pacureanu, C. Revol-Muller, J. Rose, Maria Sanchez Ruiz, F. Peyrin
Advances in imaging techniques lead to nondestructive 3D visualization of biological tissue at a sub-cellular scale. As a consequence, new demands emerge to segment complex structures. For instance, synchrotron radiation micro-CT, makes it possible to image the lacunar-canalicular porosity in bone tissue. This structure contains a dense network of slender channels interconnecting the cells. Their size (~300-600 nanometers in diameter) is at the limit of the acquisition system resolution (280 nm) making their detection difficult. In this work is proposed a variational region growing segmentation method adapted for cellular networks. To control the evolution of the segmentation through tubular structures a vesselness map is introduced in the expression of the functional to minimize. The method is tested on synthetic images and applied to experimental data.
{"title":"Vesselness-guided variational segmentation of cellular networks from 3D micro-CT","authors":"A. Pacureanu, C. Revol-Muller, J. Rose, Maria Sanchez Ruiz, F. Peyrin","doi":"10.1109/ISBI.2010.5490135","DOIUrl":"https://doi.org/10.1109/ISBI.2010.5490135","url":null,"abstract":"Advances in imaging techniques lead to nondestructive 3D visualization of biological tissue at a sub-cellular scale. As a consequence, new demands emerge to segment complex structures. For instance, synchrotron radiation micro-CT, makes it possible to image the lacunar-canalicular porosity in bone tissue. This structure contains a dense network of slender channels interconnecting the cells. Their size (~300-600 nanometers in diameter) is at the limit of the acquisition system resolution (280 nm) making their detection difficult. In this work is proposed a variational region growing segmentation method adapted for cellular networks. To control the evolution of the segmentation through tubular structures a vesselness map is introduced in the expression of the functional to minimize. The method is tested on synthetic images and applied to experimental data.","PeriodicalId":250523,"journal":{"name":"2010 IEEE International Symposium on Biomedical Imaging: From Nano to Macro","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2010-04-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128428510","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2010-04-14DOI: 10.1109/ISBI.2010.5490238
Scott Doyle, J. Monaco, A. Madabhushi, S. Lindholm, P. Ljung, Lance Ladic, J. Tomaszeweski, M. Feldman
A single digital pathology image can occupy over 10 gigabytes of hard disk space, rendering it difficult to store, analyze, and transmit. Though image compression provides a means of reducing the storage requirement, its effects on CAD (and pathologist) performance are not yet clear. In this work we assess the impact of compression on the ability of a CAD system to detect carcinoma of the prostate (CaP) in histological sections. The CAD algorithm proceeds as follows: Glands in the tissue are segmented using a region-growing algorithm. The size of each gland is then extracted and modeled using a mixture of Gamma distributions. A Markov prior (specifically, a probabilistic pairwise Markov model) is employed to encourage nearby glands to share the same class (i.e. cancerous or non-cancerous). Finally, cancerous glands are aggregated into continuous regions using a distance-hull algorithm. We evaluate CAD performance over 12 images compressed at 14 different compression ratios using JPEG2000. Algorithm performance (measured using the under the receiver operating characteristic curves) remains relatively constant for compression ratios up to 1:256. After this point performance degrades precipitously. We also have an expert pathologist view the compressed images and assign a confidence measure as to their diagnostic fidelity.
{"title":"Evaluation of effects of JPEG2000 compression on a computer-aided detection system for prostate cancer on digitized histopathology","authors":"Scott Doyle, J. Monaco, A. Madabhushi, S. Lindholm, P. Ljung, Lance Ladic, J. Tomaszeweski, M. Feldman","doi":"10.1109/ISBI.2010.5490238","DOIUrl":"https://doi.org/10.1109/ISBI.2010.5490238","url":null,"abstract":"A single digital pathology image can occupy over 10 gigabytes of hard disk space, rendering it difficult to store, analyze, and transmit. Though image compression provides a means of reducing the storage requirement, its effects on CAD (and pathologist) performance are not yet clear. In this work we assess the impact of compression on the ability of a CAD system to detect carcinoma of the prostate (CaP) in histological sections. The CAD algorithm proceeds as follows: Glands in the tissue are segmented using a region-growing algorithm. The size of each gland is then extracted and modeled using a mixture of Gamma distributions. A Markov prior (specifically, a probabilistic pairwise Markov model) is employed to encourage nearby glands to share the same class (i.e. cancerous or non-cancerous). Finally, cancerous glands are aggregated into continuous regions using a distance-hull algorithm. We evaluate CAD performance over 12 images compressed at 14 different compression ratios using JPEG2000. Algorithm performance (measured using the under the receiver operating characteristic curves) remains relatively constant for compression ratios up to 1:256. After this point performance degrades precipitously. We also have an expert pathologist view the compressed images and assign a confidence measure as to their diagnostic fidelity.","PeriodicalId":250523,"journal":{"name":"2010 IEEE International Symposium on Biomedical Imaging: From Nano to Macro","volume":"306 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2010-04-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121828436","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：