BackgroundThe extraction of overlapping cell nuclei is a critical issue in automated
diagnosis systems. Due to the similarities between overlapping and malignant
nuclei, misclassification of the overlapped regions can affect the automated
systems’ final decision. In this paper, we present a method for detecting
overlapping cell nuclei in Pap smear samples.MethodJudgement about the presence of overlapping nuclei is performed in three steps
using an unsupervised clustering approach: candidate nuclei regions are located
and refined with morphological operations; key features are extracted; and
candidate nuclei regions are clustered into two groups, overlapping or
non-overlapping, A new combination of features containing two local minima-based
and three shape-dependent features are extracted for determination of the presence
or absence of overlapping. F1 score, precision, and recall values are used to
evaluate the method’s classification performance.ResultsIn order to make evaluation, we compared the segmentation results of the
proposed system with empirical contours. Experimental results indicate that
applied morphological operations can locate most of the nuclei and produces
accurate boundaries. Independent features significance test indicates that our
feature combination is significant for overlapping nuclei. Comparisons of the
classification results of a fuzzy clustering algorithm and a non-fuzzy clustering
algorithm show that the fuzzy approach would be a more convenient mechanism for
classification of overlapping.ConclusionThe main contribution of this study is the development of a decision mechanism
for identifying overlapping nuclei to further improve the extraction process with
respect to the segmentation of interregional borders, nuclei area, and radius.
Experimental results showed that our unsupervised approach with proposed feature
combination yields acceptable performance for detection of overlapping
nuclei.
In this paper, the classification capability of Calinski-Harabasz criterion as an internal cluster validation measure has been evaluated for clustering-based region discrimination on cervical cells. In this approach, subregions in the sample image are initially randomly constructed to be the individuals of the population. At each generation, individuals are evaluated according to their Accordingly a novel genetic structure for meta heuristic area isolation is proposed. Evaluation of proposed combination of genetic algorithm and Calinski-Harabasz measure is achieved by experiments, conducted on real cervical cell samples. We have used two separate cluster validity measures to evaluate the performance of the clustering approach. Jaccard index and F-score are utilized for objective comparison. Results shows that, Calinski-Harabasz criteria may have a better performance with proposed novel genetic structure and presented mechanism may have great potential on discrimination of specific regions.
Spina bifida is a birth defect caused by incomplete closing around the spinal cord. Spina bifida is diagnosed in a number of different ways. One approach involves searching for a deformity in the spinal axis via ultrasound. Although easy to apply, this approach requires a highly trained clinician to locate the abnormality due to the noise and distortion present in prenatal ultrasound images. Accordingly, visual examination of ultrasound images may be error prone and subjective. A computerized support system that would automatically detect the location of the spinal deformity may be helpful to the clinician in the diagnostic process. Such a software system first and foremost would require an algorithm for the identification of the entire (healthy or unhealthy) spine in the ultrasound image. This paper introduces a novel flocking dynamics based approach for reducing the size of the search space in the spine identification problem. Proposed approach accepts bone-like blobs on the ultrasound images as bird flocks and combine them into bone groups by calculating the migration path of each flock. Presented results reveal that the method is able to locate correct bones to be grouped together and reduce search space (i.e. number of bones) up to 68%.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.