Statistical measurement of ultrasound placenta images using segmentation approach
01 Dec 2010-pp 309-316
TL;DR: The ultrasound screening of placenta in the initial stages of gestation helps to identify the complication induced by GDM on the placental development which accounts for the fetal growth.
Abstract: Medical diagnosis is the major challenge faced by the medical experts. Highly specialized tools are necessary to assist the experts in diagnosing the diseases. Gestational Diabetes Mellitus is a condition in pregnant women which increases the blood sugar levels. It complicates the pregnancy by affecting the placental growth. The ultrasound screening of placenta in the initial stages of gestation helps to identify the complication induced by GDM on the placental development which accounts for the fetal growth. This work focus on the classification of ultrasound placenta images into normal and abnormal images based on statistical measurements. The ultrasound images are usually low in resolution which may lead to loss of characteristic features of the ultrasound images. The placenta images obtained in an ultrasound examination is stereo mapped to reconstruct the placenta structure from the ultrasound images. The dimensionality reduction is done on stereo mapped placenta images using wavelet decomposition. The ultrasound placenta image is segmented using watershed approach to obtain the statistical measurements of the stereo mapped placenta images. Using the statistical measurements, the ultrasound placenta images are then classified as normal and abnormal using Back Propagation neural networks.
Citations
More filters
TL;DR: This review covers state‐of‐the‐art segmentation and classification methodologies for the whole fetus and, more specifically, the fetal brain, lungs, liver, heart and placenta in magnetic resonance imaging and (3D) ultrasound for the first time.
70 citations
18 Jul 2012
TL;DR: A new effective contour tracking algorithm and representation method based on the pixel vertex matrix that could effectively reduce code stream for contours and hence increase the compression ratio of the image.
Abstract: Based on analysis of contours of irregular region, and according to the characteristic that massive continuous code and the same specific code combination are usually contained in a region boundary's vertex chain code, a new effective contour tracking algorithm and representation method based on the pixel vertex matrix is proposed. Moreover, we re-encoding the new vertex chain code using a Huffman coding strategy and then select the more compressed result as the output. The results showed that the new method could effectively reduce code stream for contours, hence increase the compression ratio of the image.
1 citations
Dissertation•
13 Jul 2015
TL;DR: A new penalized likelihood iterative reconstruction algorithm for Positron Emission Tomography, based on the maximum likelihood or the least squares cost function is proposed.
Abstract: Iterative image reconstruction methods have attracted considerable attention in the past decades for applications in Computer Tomography (CT) due to the feasibility of incorporating the physical and statistical properties of the imaging process completely. So far, all statistical reconstruction algorithms are based on the maximum likelihood (ML) or the least squares cost function. The maximum likelihood-expectation maximization (ML-EM) algorithm, which is a general statistical method for seeking the estimate of the image, allows computing projections that are close to the measured projection data. Iterative based ML reconstruction algorithms require a considerable computational cost per iteration. The advantages of the iterative approach include better insensitivity to noise and capability of reconstructing an optimal image in the case of incomplete data. The method has been applied in emission tomography modalities like SPECT and PET, where there is significant attenuation along ray paths and noise statistics are relatively poor. Generally speaking, the tomography reconstruction with a limited number of data appears as a highly underdetermined ill-posed problem. The projection data generated by the CT system are initially noisy and the ML algorithm tends to increase this noise and in particular the noise artifacts through the successive iterations. This accumulation of noise leads to a premature stopping of the ML-EM reconstruction process. Several methods have been developed to decrease this accumulation of noise and improve the quality of the reconstructed images in tomography. The aim of this research is to propose a new penalized likelihood iterative reconstruction algorithm for Positron Emission Tomography, by
Additional excerpts
...Malathi et al [50] developed an algorithm to classify the ultrasound placenta images either as normal or abnormal, based on statistical measurements....
[...]
TL;DR: This work proposes a framework of dictionary-optimized sparse learning based MR super-resolution method to solve the problem of sample selection for dictionary learning of sparse reconstruction and shows that the dictionary- optimized sparse learning improves the performance of sparse representation.
Abstract: Abstract Magnetic Resonance Super-resolution Imaging Measurement (MRIM) is an effective way of measuring materials. MRIM has wide applications in physics, chemistry, biology, geology, medical and material science, especially in medical diagnosis. It is feasible to improve the resolution of MR imaging through increasing radiation intensity, but the high radiation intensity and the longtime of magnetic field harm the human body. Thus, in the practical applications the resolution of hardware imaging reaches the limitation of resolution. Software-based super-resolution technology is effective to improve the resolution of image. This work proposes a framework of dictionary-optimized sparse learning based MR super-resolution method. The framework is to solve the problem of sample selection for dictionary learning of sparse reconstruction. The textural complexity-based image quality representation is proposed to choose the optimal samples for dictionary learning. Comprehensive experiments show that the dictionary-optimized sparse learning improves the performance of sparse representation.
Additional excerpts
...The sparse representation model-based image processing performs well on image denoising [1], image deblurring [2], [3], image restoration [4]....
[...]
References
More filters
Journal Article•
TL;DR: It is concluded that diabetic's placentae showed increase in weight, central thickness and diameter as compared to normal and hypertensive group and Hypertensive's Placenta showed no significant change in shape, shape central thicknessand attachment of umbilical cord when compared with normal group.
Abstract: Background: Gestational diabetes is much more common than pre-existing diabetes i.e. it complicates 2% to 5% of pregnancies. When metabolic control is good, perinatal mortality should be no higher than in general population. However, macrosomia continuous to be a problem in higher than average proportions of such cases. Macrosomia also involves placenta within the chronic hypertensive disease, the most common diagnosis is essential vascular hypertension. Methods: Total 60 full term placenta, 20 from normal and 20 each from gestational diabetics and chronic hypertensive mothers were studied grossly. Shape, attachment of umbilical cord, weight, diameter and central thickness of all placentas were noted. Results: The study demonstrates that there is change of shape i.e. two lobes in one placenta from diabetic group. All other placentae were singly lobed and discoidal shape with central attachment of umbilical cord to the foetal surface of placenta. Weight central thickness and diameter were significantly greater in diabetic group as compared to normal and hypertensive group. Hypertensive group shows non significant decrease in weight of placentae while there was no change in central thickness and diameter of placenta in hypertensive than the normal group. Conclusions: On the basis of results of present study, it is concluded that diabetic’s placentae showed increase in weight, central thickness and diameter. One out of 20 placentae in diabetic group also showed change of shape and attachment of umbilical cord to one love. Hypertensive’s placentae showed no significant change in weight, shape central thickness and attachment of umbilical cord when compared with normal group. Key words: Placenta, Gestational diabetes, Maternal Hypertension.
56 citations
"Statistical measurement of ultrasou..." refers methods in this paper
...Using the statistical measurements, the ultrasound placenta images are then classified as normal and abnormal using Back Propagation neural networks....
[...]
01 Jan 2009
TL;DR: This paper compares the performances of the two popular region-based image segmentation methods namely the Watershed method and the Mean-shift method.
Abstract: Image segmentation and its performance evaluation are very difficult but important problems in computer vision. A major challenge in segmentation evaluation comes from the fundamental conflict between generality and objectivity: For general-purpose segmentation, the ground truth and segmentation accuracy may not be well defined, while embedding the evaluation in a specific application; the evaluation results may not be extensible to other applications. In this paper, we compare the performances of the two popular region-based image segmentation methods namely the Watershed method and the Mean-shift method. The watershed method, also called the watershed transform, is an image segmentation approach based on mathematical morphology. Mean-shift method is a data-clustering method that searches for the local maximal density points and then groups all the data to the clusters defined by these maximal density points.
40 citations
15 Apr 2004
TL;DR: The percentage of correct classification or accuracy was computed for different subsets of features, with different sizes of the region of interest showing that generally a small number of features are enough for achieving the highest accuracy.
Abstract: The characterization of ultrasonic images of the placenta is one of the clinical procedures followed for assessing the progress of pregnancy. In this work, the classification of scans of the placenta according with Grannum grading is attempted. Feature selection was used for determining the relevant textural features that were extracted from the scans. Three different sets of textural features, namely cooccurrence matrices, Laws masks and neighborhood gray-tone difference matrices (NGTDM) were used. A set of 59 images corresponding to the four grades was sampled in subimages of different sizes, the textural features were computed and weighted using the relief-F algorithm. The strategy used for classification was the k-nearest neighbor algorithm using leave-one-out cross-validation. The percentage of correct classification or accuracy was computed for different subsets of features, with different sizes of the region of interest showing that generally a small number of features are enough for achieving the highest accuracy.
18 citations
"Statistical measurement of ultrasou..." refers background in this paper
...This is accomplished by the close proximity of the maternal and fetal blood systems within the placenta [1]....
[...]
TL;DR: This paper has made an attempt to classify the placenta based on the intensity level of histogram of the ultrasound images ofPlacenta using k nearest neighbor classifier to analyze the complications of gestational diabetes mellitus on the growth of the Placenta.
Abstract: In this paper, the authors have made an attempt to classify the placenta based on the intensity level of histogram of the ultrasound images of placenta. The medical images are usually low in resolution. Specialized tools are required to assist the medical experts in medical image diagnosis and for further treatment. The image histogram is used to classify the ultrasound images of placenta into normal and abnormal placenta using k nearest neighbor classifier. It is further used to analyze the complications of gestational diabetes mellitus on the growth of the placenta.
16 citations
"Statistical measurement of ultrasou..." refers background in this paper
...Hypertensive group shows non significant decrease in weight of placenta while there was no change in central thickness and diameter of placenta in hypertensive than the normal group. it is concluded that diabetic's placenta showed increase in weight, central thickness and diameter[3]....
[...]
24 Apr 2003
TL;DR: The task of classifying ultrasonic images of the placenta according with the gradation proposed by Grannum (1979) is attempted and the ability of a decision tree classifier to discriminate different textures with three sets of textural features was tested.
Abstract: The assessment of the placenta maturity is an important issue in prenatal diagnosis. In this work, the task of classifying ultrasonic images of the placenta according with the gradation proposed by Grannum (1979) is attempted. With this purpose, the ability of a decision tree classifier to discriminate different textures with three sets of textural features was tested. The performance of the classifier using textural features corresponding to co-ocurrence matrices, Law's operators and neighborhood gray-tone difference matrices (NGDTM) was firstly assessed. A preliminary experiment was done using natural textures taken from Brodatz's (1966) album with the addition of an ultrasonic image of the placenta. In a second step the method was applied to the problem of the classification of ultrasonic images of the placenta corresponding to different grades.
11 citations
"Statistical measurement of ultrasou..." refers methods in this paper
...Using the statistical measurements, the ultrasound placenta images are then classified as normal and abnormal using Back Propagation neural networks....
[...]