An Integrated Approach for Optimal Feature Selection and Anatomic Location Identification on Pediatric Foreign Body Aspired Radiographic Images
01 Jan 2018-pp 101-128
TL;DR: This chapter mainly focuses on identifying the relative advantages of using specific combination of image enhancement, segmentation, feature extraction for optimal feature selection and proposing an approach for automatic anatomic location identification process on pediatric foreign body aspired images.
Abstract: Foreign body aspiration is a frequent pediatric emergency, with incidence peaking at two years of age. Foreign body (FB) can be described as the intrude object which is not belong to the human body. The localization of FB needs radiography X-ray, CT, MRI assessment. Foreign bodies such as coin and metallic items are easily seen on radiographs, but it is difficult to identify food and plastic objects on foreign body aspired radiography images. The process of location identification takes more time which leads more complication, even it leads to fatal (Lecron and Benjelloun in Med Imaging SPIE Proc 8314:1–8, 2012 [8]). Therefore, the proposed work aims to develop an approach for identifying the anatomic location in which the complications of diagnosis process will be reduced. Image processing plays vital role in this scenario, especially in automating the process of determining the anatomic location of the foreign body on pediatric radiographic images. This chapter mainly focuses on identifying the relative advantages of using specific combination of image enhancement, segmentation, feature extraction for optimal feature selection and proposing an approach for automatic anatomic location identification process on pediatric foreign body aspired images. This process includes the radiographic image acquisition of the foreign body aspired pediatric patients, image enhancement, and segmentation methods. The identification of suitable segmentation method for extracting the optimal features is related to a range of research studies published on image segmentation, feature extraction, feature selection methods. The observation of the existing work helps to understand the importance of various segmentation methods and also supports to develop improved segmentation methods such as constraint-based median filtering, constraint-based iterative thresholding, constraint-based Sobel boundary detection, and K-means clustering. The ability of the enhanced segmentation techniques are determined by the performance comparison with the existing segmentation techniques which is done by the quality metrics evaluation. The feature extraction is used for describing the true region of interest based on shape-, edge-, and texture-based descriptors. The most influenced features are identified by applying hybrid feature selection method which is a combination of filter and wrapper methods in predicting the location and shape of the foreign body. A novel, automatic anatomic location identification approach (AALIA) using 8-connected block searching algorithm and corner identification methods are applied to identification and classification of the anatomic location of foreign body. To evaluate performance of the proposed approach, the accuracy measure precision, recall, F-Measure, and receiver operator characteristic (ROC) with respect to sensitivity, specificity, positive predicted rate, and negative predicted rate are considered. The results obtained from the developed approach are comparatively better than the existing works.
References
More filters
TL;DR: The contributions of this special issue cover a wide range of aspects of variable selection: providing a better definition of the objective function, feature construction, feature ranking, multivariate feature selection, efficient search methods, and feature validity assessment methods.
Abstract: Variable and feature selection have become the focus of much research in areas of application for which datasets with tens or hundreds of thousands of variables are available. These areas include text processing of internet documents, gene expression array analysis, and combinatorial chemistry. The objective of variable selection is three-fold: improving the prediction performance of the predictors, providing faster and more cost-effective predictors, and providing a better understanding of the underlying process that generated the data. The contributions of this special issue cover a wide range of aspects of such problems: providing a better definition of the objective function, feature construction, feature ranking, multivariate feature selection, efficient search methods, and feature validity assessment methods.
14,509 citations
01 Jan 1997
TL;DR: A survey of machine learning methods for handling data sets containing large amounts of irrelevant information can be found in this article, where the authors focus on two key issues: selecting relevant features and selecting relevant examples.
Abstract: In this survey, we review work in machine learning on methods for handling data sets containing large amounts of irrelevant information. We focus on two key issues: the problem of selecting relevant features, and the problem of selecting relevant examples. We describe the advances that have been made on these topics in both empirical and theoretical work in machine learning, and we present a general framework that we use to compare different methods. We close with some challenges for future work in this area. @ 1997 Elsevier Science B.V.
2,947 citations
TL;DR: This survey reviews work in machine learning on methods for handling data sets containing large amounts of irrelevant information and describes the advances that have been made in both empirical and theoretical work in this area.
2,869 citations
University of Colorado Denver1, Medical College of Wisconsin2, Cincinnati Children's Hospital Medical Center3, Harvard University4, Loma Linda University5, University of Louisville6, University of Arkansas7, North Shore-LIJ Health System8, Boston Children's Hospital9, Medical University of South Carolina10, Baylor College of Medicine11, University of Pennsylvania12
TL;DR: This guideline is intended as an educational tool that may help inform pediatric endoscopists in managing foreign body ingestions in children and may be revised as needed to account for new data, changes in clinical practice, or availability of new technology.
Abstract: Foreign body ingestions in children are some of the most challenging clinical scenarios facing pediatric gastroenterologists. Determining the indications and timing for intervention requires assessment of patient size, type of object ingested, location, clinical symptoms, time since ingestion, and myriad other factors. Often the easiest and least anxiety-producing decision is the one to proceed to endoscopic removal, instead of observation alone. Because of variability in pediatric patient size, there are less firm guidelines available to determine which type of object will safely pass, as opposed to the clearer guidelines in the adult population. In addition, the imprecise nature of the histories often leaves the clinician to question the timing and nature of the ingestion. Furthermore, changes in the types of ingestions encountered, specifically button batteries and high-powered magnet ingestions, create an even greater potential for severe morbidity and mortality among children. As a result, clinical guidelines regarding management of these ingestions in children remain varied and sporadic, with little in the way of prospective data to guide their development. An expert panel of pediatric endoscopists was convened and produced the present article that outlines practical clinical approaches to the pediatric patient with a variety of foreign body ingestions. This guideline is intended as an educational tool that may help inform pediatric endoscopists in managing foreign body ingestions in children. Medical decision making, however, remains a complex process requiring integration of clinical data beyond the scope of these guidelines. These guidelines should therefore not be considered to be a rule or to be establishing a legal standard of care. Caregivers may well choose a course of action outside of those represented in these guidelines because of specific patient circumstances. Furthermore, additional clinical studies may be necessary to clarify aspects based on expert opinion instead of published data. Thus, these guidelines may be revised as needed to account for new data, changes in clinical practice, or availability of new technology.
407 citations
TL;DR: A comprehensive solution for automatically detecting, identifying, and segmenting vertebrae in CT images by designing a framework that takes an arbitrary CT image as input and provides a segmentation in form of labelled triangulated vertebra surface models.
346 citations