Segmentation of anatomical structures in chest radiographs using supervised methods: a comparative study on a public database

Abstract

The task of segmenting the lung fields, the heart, and the clavicles in standard posterior-anterior chest radiographs is considered. Three supervised segmentation methods are compared: active shape models, active appearance models and a multi-resolution pixel classification method that employs a multi-scale filter bank of Gaussian derivatives and a k-nearest-neighbors classifier. The methods have been tested on a publicly available database of 247 chest radiographs, in which all objects have been manually segmented by two human observers. A parameter optimization for active shape models is presented, and it is shown that this optimization improves performance significantly. It is demonstrated that the standard active appearance model scheme performs poorly, but large improvements can be obtained by including areas outside the objects into the model. For lung field segmentation, all methods perform well, with pixel classification giving the best results: a paired t-test showed no significant performance difference between pixel classification and an independent human observer. For heart segmentation, all methods perform comparably, but significantly worse than a human observer. Clavicle segmentation is a hard problem for all methods; best results are obtained with active shape models, but human performance is substantially better. In addition, several hybrid systems are investigated. For heart segmentation, where the separate systems perform comparably, significantly better performance can be obtained by combining the results with majority voting. As an application, the cardio-thoracic ratio is computed automatically from the segmentation results. Bland and Altman plots indicate that all methods perform well when compared to the gold standard, with confidence intervals from pixel classification and active appearance modeling very close to those of a human observer. All results, including the manual segmentations, have been made publicly available to facilitate future comparative studies.

Global and Local Multi-valued Dissimilarity-Based Classification: Application to Computer-Aided Detection of Tuberculosis

“…Prior to feature extraction, lung fields were automatically segmented from the radiographs using multi-resolution pixel classification, with settings as given in [5]. In order to train this segmentation procedure, lung fields were segmented manually from 20 radiographs not used otherwise in this study.…”

Visual detection of knives in security applications using Active Appearance Models

“…Due to their generic nature and high effectiveness in locating objects, numerous applications in medical image interpretation followed. The typical medical application is finding an object, usually an organ in a medical image of a specific body part, such as locating the bone in a magnetic resonance image (MRI) of the knee [5], the left and right ventricles in a cardiac MRI [18], or the heart in a chest radiograph [24].…”

Computational methods for the analysis of functional 4D-CT chest images.

“…To address the shape variability for each subject, the shape is adapted for the segmentation of further time-point images with the previously segmented images from the same subject. Van Ginneken et al [52] optimized the active shape model (ASM) developed by Tsai et al [53] to segment the lung fields. They compared the segmentation with an active appearance model (AAM)-based segmentation and multi-scale resolution pixel classification, concluding that the latter gives the best results.…”

“…They compared the segmentation with an active appearance model (AAM)-based segmentation and multi-scale resolution pixel classification, concluding that the latter gives the best results. Hardie et al [54] invoked the optimized ASM shape model of van Ginneken et al [52] to segment the lungs field in a CAD system developed to identify lung nodules on CT images. Sun et al [55] segmented the lungs in two main processing steps.…”

“…In the application to CT chest images, most of the earlier segmentation techniques [32,35,47,52,[65][66][67] presume only the lungs are darker than the other chest tissues, which might result in failure to detect nodules in the case of severe lung pathologies. To avoid such failures, more recent lung segmentation methods, which will be briefly reviewed in this section, consider visual appearances [1,357], shapes [55,[358][359][360][361][362], or hybrid techniques [60,62,[363][364][365] to account for normal and pathological tissues.…”