Segmenting Multiple Objects with Overlapping Appearance and Uncertainty (original) (raw)
Related papers
Probabilistic Segmentation of the Knee Joint from X-ray Images
A probabilistic method is proposed for segmentation of the knee joint. A likelihood function is formulated that explicitly models overlapping object appearance. Priors on global appearance and geometry (including shape) are learned from example images. Markov chain Monte Carlo methods are used to obtain samples from a posterior distribution over model parameters from which expectations can be estimated. The result is a probabilistic segmen- tation that quantifies uncertainty so that measurements such as joint space can be made with associated uncertainty. Joint space area and mean point-to-contour distance are used for evaluation. The aim of this paper is to outline a probabilistic, model-based segmentation method for the knee joint from x-ray images and to make explicit the uncertainty in the segmentation so obtained. The method explicitly handles the possible overlapping of femur and tibia and their appearance models. Such cases are not handled in methods based on active contours ...
Bayesian inference for model-based segmentation of computed radiographs of the hand
Artificial Intelligence in Medicine, 1993
We present a method for medical image understanding by computer that uses model-based, hierarchical Bayesian inference to accurately segment imaged anatomy. A first application is a prototype system that automatically segments and measures symptoms of arthridities in hand radiographs. This is potentially useful in radiological diagnosis and tracking of arthridities. Key steps of the model-based, Bayesian inference approach are: (1) prediction of imagery features from 3D models of anatomy, pammeterized by population statistics, (2) local image feature extraction in predicted sub-regions, and (3) the use of a probabilistic calculus to accrue results of image processing and image feature matching procedures in support or denial of hypothesesabout the imaged anatomy. The prototype system for hand mdiograph analysisaccurately segments normal and somewhat degenerated hand anatomy. Results are shown of the ability of the automated system to 'fail soft', recognizing when segmentation is inadequate for accurate measurement. This self evaluation capability improves reliability of measurements for potential clinical use.
Probabilistic Model Based Image Segmentation
The International journal of Multimedia & Its Applications, 2014
There exists a plethora of algorithms to perform image segmentation and there are several issues related to execution time of these algorithms. Image Segmentation is nothing but label relabeling problem under probability framework. To estimate the label configuration, an iterative optimization scheme is implemented to alternately carry out the maximum a posteriori (MAP) estimation and the maximum likelihood (ML) estimations. In this paper this technique is modified in such a way so that it performs segmentation within stipulated time period. The extensive experiments shows that the results obtained are comparable with existing algorithms. This algorithm performs faster execution than the existing algorithm to give automatic segmentation without any human intervention. Its result match image edges very closer to human perception.
An articulated statistical shape model for accurate hip joint segmentation
Conference proceedings : ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual Conference, 2009
In this paper we propose a framework for fully automatic, robust and accurate segmentation of the human pelvis and proximal femur in CT data. We propose a composite statistical shape model of femur and pelvis with a flexible hip joint, for which we extend the common definition of statistical shape models as well as the common strategy for their adaptation. We do not analyze the joint flexibility statistically, but model it explicitly by rotational parameters describing the bent in a ball-and-socket joint. A leave-one-out evaluation on 50 CT volumes shows that image driven adaptation of our composite shape model robustly produces accurate segmentations of both proximal femur and pelvis. As a second contribution, we evaluate a fine grain multi-object segmentation method based on graph optimization. It relies on accurate initializations of femur and pelvis, which our composite shape model can generate. Simultaneous optimization of both femur and pelvis yields more accurate results than...
Probabilistic models for robot-based object segmentation
Robotics and Autonomous Systems, 2011
This paper introduces a novel probabilistic method for robot based object segmentation. The method integrates knowledge of the robot's motion to determine the shape and location of objects. This allows a robot with no prior knowledge of its workspace to isolate objects against their surroundings by moving them and observing their visual feedback. The main contribution of the paper is to improve upon current methods by allowing object segmentation in changing environments and moving backgrounds. The approach allows optimal values for the algorithm parameters to be estimated. Empirical studies against alternatives demonstrate clear improvements in both planar and three dimensional motion.
Bone segmentation in CT-angiography data using a probabilistic atlas
Proc. VMV …, 2003
Automatic segmentation of bony structures in CT angiography datasets is an essential pre-processing step necessary for most visualization and analysis tasks. Since traditional density and gradient operators fail in non-trivial cases (or at last require extensive operator work), we propose a new method for segmentation of CTA data based on a probabilistic atlas. Storing densities and masks of previously manually segmented tissues to the atlas can constitute a statistical information base for latter accurate segmentation. In order to eliminate dimensional and anatomic variability of the atlas input datasets, these have to be spatially normalized (registered) first by applying a non-rigid transformation. After this transformation, densities and tissue masks are statistically processed (e.g. averaged) within the atlas. Records in the atlas can be later evaluated for estimating the probability of bone tissue in a voxel of an unsegmented dataset.
Automatic Extraction of Femur Contours from Calibrated X-Ray Images using Statistical Information
Journal of Multimedia, 2007
Automatic identification and extraction of bone contours from x-ray images is an essential first step task for further medical image analysis. In this paper we propose a 3D statistical model based framework for the proximal femur contour extraction from calibrated x-ray images. The automatic initialization to align the 3D model with the x-ray images is solved by an Estimation of Bayesian Network Algorithm to fit a simplified multiple component geometrical model of the proximal femur to the x-ray data. Landmarks can be extracted from the geometrical model for the initialization of the 3D statistical model. The contour extraction is then accomplished by a joint registration and segmentation procedure. We iteratively updates the extracted bone contours and an instanced 3D model to fit the x-ray images. Taking the projected silhouettes of the instanced 3D model on the registered x-ray images as templates, bone contours can be extracted by a graphical model based Bayesian inference. The 3D model can then be updated by a non-rigid 2D/3D registration between the 3D statistical model and the extracted bone contours. Preliminary experiments on clinical data sets verified its validity.
Markov surfaces: A probabilistic framework for user-assisted three-dimensional image segmentation
Computer Vision and Image Understanding, 2011
This paper presents Markov surfaces, a probabilistic algorithm for user-assisted segmentation of elongated structures in 3D images. The 3D segmentation problem is formulated as a path-finding problem, where path probabilities are described by Markov chains. Users define points, curves, or regions on 2D image slices, and the algorithm connects these user-defined features in a way that respects the underlying elongated structure in data. Transition probabilities in the Markov model are derived from intensity matches and interslice correspondences, which are generated from a slice-to-slice registration algorithm. Bezier interpolations between paths are applied to generate smooth surfaces. Subgrid accuracy is achieved by linear interpolations of image intensities and the interslice correspondences. Experimental results on synthetic and real data demonstrate that Markov surfaces can segment regions that are defined by texture, nearby context, and motion. A parallel implementation on a streaming parallel computer architecture, a graphics processor, makes the method interactive for 3D data.
Probabilistic multi-shape segmentation of knee extensor and flexor muscles
Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention, 2011
Patients with chronic obstructive pulmonary disease (COPD) often exhibit skeletal muscle weakness in lower limbs. Analysis of the shapes and sizes of these muscles can lead to more effective therapy. Unfortunately, segmenting these muscles from one another is a challenging task due to a lack of image information in many areas. We present a fully automatic segmentation method that overcomes the inherent difficulties of this problem to accurately segment the different muscles. Our method enforces a multi-region shape prior on the segmentation to ensure feasibility and provides an energy minimizing probabilistic segmentation that indicates areas of uncertainty. Our experiments on 3D MRI datasets yield an average Dice similarity coefficient of 0.92 +/- 0.03 with the ground truth.
2008
Local graph-based probabilistic representation of object shape and appearance for model-based medical image segmentation Image segmentation is the process of partitioning a digital image into regions originating from different objects in the scene. The segmentation of anatomical objects is indispensable for the analysis of medical images. It enables the assessment of anatomical measurements and it is a possible means towards diagnosis, therapy planning and visualization. As anatomical objects appear in medical images with high variability, the construction of a model that incorporates prior knowledge about these objects is essential during segmentation. A popular and very effective approach is to represent the shape as a set of landmark points, and learn the shape variations from a set of example shapes. Whereas conventional methods build a global point distribution model that considers correlations between all points in the set, this thesis presents a localized model that captures statistical prior shape information as a concatenation of multiple local shape models into a deformable graph configuration. The method has a strong theoretical basis as the model construction and model fitting are formulated from a probability point of view. Its validity and highly generic nature are illustrated for the segmentation of multiple anatomical structures, both from two-as threedimensional images. A comparison to methods that use a global model shows that the presented approach, thanks to its localized nature, is able to fit more accurately to unseen objects.