Method for segmenting chest CT image data using an anatomical model: preliminary results

Automated segmentation of lungs with severe interstitial lung disease in CT

PURPOSE

Accurate segmentation of lungs with severe interstitial lung disease (ILD) in thoracic computed tomography (CT) is an important and difficult task in the development of computer-aided diagnosis (CAD) systems. Therefore, we developed in this study a texture analysis-based method for accurate segmentation of lungs with severe ILD in multidetector CT scans.

METHODS

Our database consisted of 76 CT scans, including 31 normal cases and 45 abnormal cases with moderate or severe ILD. The lungs in three selected slices for each CT scan were first manually delineated by a medical physicist, and then confirmed or revised by an expert chest radiologist, and they were used as the reference standard for lung segmentation. To segment the lungs, we first employed a CT value thresholding technique to obtain an initial lung estimate, including normal and mild ILD lung parenchyma. We then used texture-feature images derived from the co-occurrence matrix to further identify abnormal lung regions with severe ILD. Finally, we combined the identified abnormal lung regions with the initial lungs to generate the final lung segmentation result. The overlap rate, volume agreement, mean absolute distance (MAD), and maximum absolute distance (dmax) between the automatically segmented lungs and the reference lungs were employed to evaluate the performance of the segmentation method.

RESULTS

Our segmentation method achieved a mean overlap rate of 96.7%, a mean volume agreement of 98.5%, a mean MAD of 0.84 mm, and a mean dmax of 10.84 mm for all the cases in our database; a mean overlap rate of 97.7%, a mean volume agreement of 99.0%, a mean MAD of 0.66 mm, and a mean dmax of 9.59 mm for the 31 normal cases; and a mean overlap rate of 96.1%, a mean volume agreement of 98.1%, a mean MAD of 0.96 mm, and a mean dmax of 11.71 mm for the 45 abnormal cases with ILD.

CONCLUSIONS

Our lung segmentation method provided accurate segmentation results for abnormal CT scans with severe ILD and would be useful for developing CAD systems for quantification, detection, and diagnosis of ILD.

Update on radiology of emphysema and therapeutic implications

Emerging treatments require appropriate CT targeting of a selected lobe or lobes and target airways to obtain a successful response. CT scan is used in pretreatment planning to select patients and plan treatment strategy and posttreatment to confirm correct deployment of devices and assess treatment response. Increasingly treatments are being developed to treat patients who have emphysema who require accurate quantitation of extent and distribution of the process. Functional assessment can be made by inference of detailed anatomic correlates and by direct measurement of regional function using dynamic scan protocols. This article summarizes the current role of imaging in the assessment of patients who have emphysema.

Texture classification-based segmentation of lung affected by interstitial pneumonia in high-resolution CT

Accurate and automated lung field (LF) segmentation in high-resolution computed tomography (HRCT) is highly challenged by the presence of pathologies affecting lung borders, also affecting the performance of computer-aided diagnosis (CAD) schemes. In this work, a two-dimensional LF segmentation algorithm adapted to interstitial pneumonia (IP) patterns is presented. The algorithm employs k-means clustering followed by a filling operation to obtain an initial LF order estimate. The final LF border is obtained by an iterative support vector machine neighborhood labeling of border pixels based on gray level and wavelet coefficient statistics features. A second feature set based on gray level averaging and gradient features was also investigated to evaluate its effect on segmentation performance of the proposed method. The proposed method is evaluated on a dataset of 22 HRCT cases spanning a range of IP patterns such as ground glass, reticular, and honeycombing. The accuracy of the method is assessed using area overlap and shape differentiation metrics (d(mean), d(rms), and d(max)), by comparing automatically derived lung borders to manually traced ones, and further compared to a gray level thresholding-based (GLT-based) method. Accuracy of the methods evaluated is also compared to interobserver variability. The proposed method incorporating gray level and wavelet coefficient statistics demonstrated the highest segmentation accuracy, averaged over left and right LFs (overlap=0.954, d(mean)=1.080 mm, d(rms)=1.407 mm, and d(max)=4.944 mm), which is statistically significant (two-tailed student's t test for paired data, p<0.0083) with respect to all metrics considered as compared to the proposed method incorporating gray level averaging and gradient features (overlap=0.918, d(mean)=2.354 mm, d(rms)=3.711 mm, and d(max)=14.412 mm) and the GLT-based method (overlap=0.897, d(mean)=3.618 mm, d(rms)=5.007 mm, and d(max)=16.893 mm). The performance of the three segmentation methods, although decreased as IP pattern severity level (mild, moderate, and severe) was increased, did not demonstrate statistically significant difference (two-tailed student's t test for unpaired data, p>0.0167 for all metrics considered). Finally, the accuracy of the proposed method, based on gray level and wavelet coefficient statistics ranges within interobserver variability. The proposed segmentation method could be used as an initial stage of a CAD scheme for IP patterns.

Automatic segmentation of lung parenchyma in the presence of diseases based on curvature of ribs

RATIONALE AND OBJECTIVES

Segmentation of lungs using high-resolution computer tomographic images in the setting of diffuse lung diseases is a major challenge in medical image analysis. Threshold-based techniques tend to leave out lung regions that have increased attenuation, such as in the presence of interstitial lung disease. In contrast, streak artifacts can cause the lung segmentation to "leak" into the chest wall. The purpose of this work was to perform segmentation of the lungs using a technique that selects an optimal threshold for a given patient by comparing the curvature of the lung boundary to that of the ribs.

METHODS

Our automated technique goes beyond fixed threshold-based approaches to include lung boundary curvature features. One would expect the curvature of the ribs and the curvature of the lung boundary around the ribs to be very close. Initially, the ribs are segmented by applying a threshold algorithm followed by morphologic operations. The lung segmentation scheme uses a multithreshold iterative approach. The threshold value is verified until the curvature of the ribs and the curvature of the lung boundary are closely matched. The curve of the ribs is represented using polynomial interpolation, and the lung boundary is matched in such a way that there is minimal deviation from this representation. Performance of this technique was compared with conventional (fixed threshold) lung segmentation techniques on 25 subjects using a volumetric overlap fraction measure.

RESULTS

The performance of the rib segmentation technique was significantly different from conventional techniques with an average higher mean volumetric overlap fraction of about 5%.

CONCLUSIONS

The technique described here allows for accurate quantification of volumetric computed tomography and more advanced segmentation of abnormal areas.

Nodule detection in a lung region that's segmented with using genetic cellular neural networks and 3D template matching with fuzzy rule based thresholding

Objective

The purpose of this study was to develop a new method for automated lung nodule detection in serial section CT images with using the characteristics of the 3D appearance of the nodules that distinguish themselves from the vessels.

Materials and Methods

Lung nodules were detected in four steps. First, to reduce the number of region of interests (ROIs) and the computation time, the lung regions of the CTs were segmented using Genetic Cellular Neural Networks (G-CNN). Then, for each lung region, ROIs were specified with using the 8 directional search; +1 or -1 values were assigned to each voxel. The 3D ROI image was obtained by combining all the 2-Dimensional (2D) ROI images. A 3D template was created to find the nodule-like structures on the 3D ROI image. Convolution of the 3D ROI image with the proposed template strengthens the shapes that are similar to those of the template and it weakens the other ones. Finally, fuzzy rule based thresholding was applied and the ROI's were found. To test the system's efficiency, we used 16 cases with a total of 425 slices, which were taken from the Lung Image Database Consortium (LIDC) dataset.

Results

The computer aided diagnosis (CAD) system achieved 100% sensitivity with 13.375 FPs per case when the nodule thickness was greater than or equal to 5.625 mm.

Conclusion

Our results indicate that the detection performance of our algorithm is satisfactory, and this may well improve the performance of computer-aided detection of lung nodules.

Adaptive border marching algorithm: automatic lung segmentation on chest CT images

Segmentation of the lungs in chest-computed tomography (CT) is often performed as a preprocessing step in lung imaging. This task is complicated especially in presence of disease. This paper presents a lung segmentation algorithm called adaptive border marching (ABM). Its novelty lies in the fact that it smoothes the lung border in a geometric way and can be used to reliably include juxtapleural nodules while minimizing oversegmentation of adjacent regions such as the abdomen and mediastinum. Our experiments using 20 datasets demonstrate that this computational geometry algorithm can re-include all juxtapleural nodules and achieve an average oversegmentation ratio of 0.43% and an average under-segmentation ratio of 1.63% relative to an expert determined reference standard. The segmentation time of a typical case is under 1min on a typical PC. As compared to other available methods, ABM is more robust, more efficient and more straightforward to implement, and once the chest CT images are input, there is no further interaction needed from users. The clinical impact of this method is in potentially avoiding false negative CAD findings due to juxtapleural nodules and improving volumetry and doubling time accuracy.

Investigation of hyperpolarized 3He magnetic resonance imaging utility in examining human airway diameter behavior in asthma through comparison with high-resolution computed tomography

RATIONALE AND OBJECTIVES

Application of a previously developed model-based algorithm on hyperpolarized (HP) (3)He magnetic resonance (MR) dynamic projection images of phantoms was extended to investigate the utility of HP (3)He MR imaging (MRI) in quantifying airway caliber changes associated with asthma.

MATERIALS AND METHODS

Airways of seven volunteers were imaged and measured using HP (3)He MRI and multidetector-row computed tomography (MDCT) before and after a methacholine (MCh) challenge. MDCT data were obtained at functional residual capacity and 1 L above functional residual capacity.

RESULTS

Comparison of the resultant data showed that HP (3)He MRI did not match MDCT in measuring the ratios of airway calibers before and after the MCh challenge in 37% to 43% of the airways from the first six generations at the two lung volumes tested. However, MDCT did yield the observation that 49% to 69% of these airways displayed bronchodilation following MCh challenge.

CONCLUSION

The current implementation of HP (3)He MRI did not match the MCh-induced postchallenge-to-prechallenge airway caliber ratios as measured with MDCT. Elevated parenchymal tethering due to bronchoconstriction-induced hyperinflation was proposed as a possible explanation for this airway dilation.

Automatic segmentation of the lungs using robust level sets

This paper presents a method for the automatic segmentation of the lungs in X-ray computed tomography (CT) images. The proposed technique is based on the use of a robust geometric active contour that is initialized around the lungs, automatically splits in two, and performs outlier rejection during the curve evolution. The technique starts by grey-level thresholding of the images followed by edge detection. Then the edge connected points are organized into strokes and classified as valid or invalid. A confidence degree (weight) is assigned to each stroke and updated during the evolution process with the valid strokes receiving a high confidence degree and the confidence degrees of the outlier strokes tending to zero. These weights depend on the distance between the stroke points and the curve and also on the stroke size. Initialization of the curve is fully automatic. Experimental results show the effectiveness of the proposed technique.

Combining 2D wavelet edge highlighting and 3D thresholding for lung segmentation in thin-slice CT

The first step in lung analysis by CT is the identification of the lung border. To deal with the increased number of sections per scan in thin-slice multidetector CT, it has been crucial to develop accurate and automated lung segmentation algorithms. In this study, an automated method for lung segmentation of thin-slice CT data is presented. The method exploits the advantages of a two-dimensional wavelet edge-highlighting step in lung border delineation. Lung volume segmentation is achieved with three-dimensional (3D) grey level thresholding, using a minimum error technique. 3D thresholding, combined with the wavelet pre-processing step, successfully deals with lung border segmentation challenges, such as anterior or posterior junction lines and juxtapleural nodules. Finally, to deal with mediastinum border under-segmentation, 3D morphological closing with a spherical structural element is applied. The performance of the proposed method is quantitatively assessed on a dataset originating from the Lung Imaging Database Consortium (LIDC) by comparing automatically derived borders with the manually traced ones. Segmentation performance, averaged over left and right lung volumes, for lung volume overlap is 0.983+/-0.008, whereas for shape differentiation in terms of mean distance it is 0.770+/-0.251 mm (root mean square distance is 0.520+/-0.008 mm; maximum distance is 3.327+/-1.637 mm). The effect of the wavelet pre-processing step was assessed by comparing the proposed method with the 3D thresholding technique (applied on original volume data). This yielded statistically significant differences for all segmentation metrics (p<0.01). Results demonstrate an accurate method that could be used as a first step in computer lung analysis by CT.

Automated classification of lung bronchovascular anatomy in CT using AdaBoost

Lung CAD systems require the ability to classify a variety of pulmonary structures as part of the diagnostic process. The purpose of this work was to develop a methodology for fully automated voxel-by-voxel classification of airways, fissures, nodules, and vessels from chest CT images using a single feature set and classification method. Twenty-nine thin section CT scans were obtained from the Lung Image Database Consortium (LIDC). Multiple radiologists labeled voxels corresponding to the following structures: airways (trachea to 6th generation), major and minor lobar fissures, nodules, and vessels (hilum to peripheral), and normal lung parenchyma. The labeled data was used in conjunction with a supervised machine learning approach (AdaBoost) to train a set of ensemble classifiers. Each ensemble classifier was trained to detect voxels part of a specific structure (either airway, fissure, nodule, vessel, or parenchyma). The feature set consisted of voxel attenuation and a small number of features based on the eigenvalues of the Hessian matrix (used to differentiate structures by shape). When each ensemble classifier was composed of 20 weak classifiers, the AUC values for the airway, fissure, nodule, vessel, and parenchyma classifiers were 0.984+/-0.011, 0.949+/-0.009, 0.945+/-0.018, 0.953+/-0.016, and 0.931+/-0.015, respectively. The strong results suggest that this could be an effective input to higher-level anatomical based segmentation models with the potential to improve CAD performance.

CAD in clinical trials: Current role and architectural requirements

Computer-aided diagnosis (CAD) technology is becoming an important tool to assess treatment response in clinical trials. However, CAD software alone is not sufficient to conduct an imaging-based clinical trial. There are a number of architectural requirements such as image receive (from multiple field sites), a database for storing quantitative measures, and data mining and reporting capabilities. In this paper we describe the architectural requirements to incorporate CAD into clinical trials and illustrate their functionality in therapeutic trials for emphysema.

Automated lung segmentation of diseased and artifact-corrupted magnetic resonance sections

Segmentation of the lungs within magnetic resonance (MR) scans is a necessary step in the computer-based analysis of thoracic MR images. This process is often confounded by image acquisition artifacts and disease-induced morphological deformation. We have developed an automated method for lung segmentation that is insensitive to these complications. The automated method was applied to 23 thoracic MR scans (413 sections) obtained from 10 patients. Two radiologists manually outlined the lung regions in a random sample of 101 sections (n=202 lungs), and the extent to which disease or artifact confounded lung border visualization was evaluated. Accuracy of lung regions extracted by the automated segmentation method was quantified by comparison with the radiologist-defined lung regions using an area overlap measure (AOM) that ranged from 0 (disjoint lung regions) to 1 (complete overlap). The AOM between each observer and the automated method was 0.82 when averaged over all lungs. The average AOM in the lung bases, where lung segmentation is most difficult, was 0.73.

A KNOWLEDGE-BASED LUNG NODULE DETECTION SYSTEM FOR HELICAL CT IMAGES

In this paper, we describe a knowledge-based system for segmenting and labeling lung nodule on helical CT images. The system was developed under a blackboard environment that incorporates a lung knowledge model, image processing model, inference engine and a blackboard. Lung model, which contains both analogical and propositional knowledge about lung in the form of semantic networks, was used to guide the interpretation process. The system works in a hierarchical structure, from large structures to the final nodule candidates by focusing on the interested region step by step. The symbolic variables, introduced to accomplish high-level inference, were defined by fuzzy confidence functions in the lung model. Composite fuzzy functions were applied to evaluate the plausibility of the mapping between the image and lung model objects. Anatomical lung segments knowledge was embedded in the system to direct 3D validation of suspicious objects. Structures were identified and abnormal objects were reported. The experimental results obtained demonstrate the proof of concept and the potential of the automated knowledge-based lung nodule detection system.

Automatic segmentation and recognition of anatomical lung structures from high-resolution chest CT images

This paper describes a fully automated segmentation and recognition scheme, which is designed to recognize lung anatomical structures in the human chest by segmenting the different chest internal organ and tissue regions sequentially from high-resolution chest CT images. A sequential region-splitting process is used to segment lungs, airway of bronchus, lung lobes and fissures based on the anatomical structures and statistical intensity distributions in CT images. The performance of our scheme is evaluated by segmenting lung structures from high-resolution multi-slice chest CT images from 44 patients; the validity of our method was proved by preliminary experimental results.

3D computerized segmentation of lung volume with computed tomography

Three-dimensional (3D)-based detection and diagnosis has an important role for significantly improving the detection and diagnosis of lung cancer upon computed tomography (CT). This report presents a 3D-based method for segmenting and visualizing lung volume by using CT images. An anisotropic filtering method was developed on CT slices to enhance the signal-to-noise ratio, and a wavelet transform-based interpolation method was used combined with volume rendering to construct the 3D volumetric data based on entire CT slices. Then an adaptive 3D region-growing algorithm was designed to segment lung volume, incorporated by automatic seed-locating methods through fuzzy logic algorithms and 3D morphological closing approaches. In addition, a 3D visualization tool was designed to view volumetric data, projections, or intersections of the lung volume at any view angle. This segmentation method was tested on single-detector CT images by percentage of volume overlap and percentage of volume difference. The experiment results show that the developed 3D-based segmentation method is effective and robust. This study lays the groundwork for 3D-based computerized detection and diagnosis of lung cancer with CT imaging. In addition, this approach can be integrated into a picture archiving and communication system serving as a visualization tool for radiologists' reading and interpretation.

Computer analysis of computed tomography scans of the lung: a survey

Current computed tomography (CT) technology allows for near isotropic, submillimeter resolution acquisition of the complete chest in a single breath hold. These thin-slice chest scans have become indispensable in thoracic radiology, but have also substantially increased the data load for radiologists. Automating the analysis of such data is, therefore, a necessity and this has created a rapidly developing research area in medical imaging. This paper presents a review of the literature on computer analysis of the lungs in CT scans and addresses segmentation of various pulmonary structures, registration of chest scans, and applications aimed at detection, classification and quantification of chest abnormalities. In addition, research trends and challenges are identified and directions for future research are discussed.

Atlas-driven lung lobe segmentation in volumetric X-ray CT images

High-resolution X-ray computed tomography (CT) imaging is routinely used for clinical pulmonary applications. Since lung function varies regionally and because pulmonary disease is usually not uniformly distributed in the lungs, it is useful to study the lungs on a lobe-by-lobe basis. Thus, it is important to segment not only the lungs, but the lobar fissures as well. In this paper, we demonstrate the use of an anatomic pulmonary atlas, encoded with a priori information on the pulmonary anatomy, to automatically segment the oblique lobar fissures. Sixteen volumetric CT scans from 16 subjects are used to construct the pulmonary atlas. A ridgeness measure is applied to the original CT images to enhance the fissure contrast. Fissure detection is accomplished in two stages: an initial fissure search and a final fissure search. A fuzzy reasoning system is used in the fissure search to analyze information from three sources: the image intensity, an anatomic smoothness constraint, and the atlas-based search initialization. Our method has been tested on 22 volumetric thin-slice CT scans from 12 subjects, and the results are compared to manual tracings. Averaged across all 22 data sets, the RMS error between the automatically segmented and manually segmented fissures is 1.96 +/- 0.71 mm and the mean of the similarity indices between the manually defined and computer-defined lobe regions is 0.988. The results indicate a strong agreement between the automatic and manual lobe segmentations.

Automatic segmentation of the lungs using multiple active contours and outlier model

This paper presents a method for the automatic segmentation of the lungs in X-ray computed tomography (CT) images. The proposed technique is based on the use of multiple active contour models (ACMs) for the simultaneous segmentation of both lungs and outlier detection. The technique starts by grey-level thresholding of the images followed by edge detection. Then the edge points are organized in strokes and a set of weights summing to one is assigned to each stroke. These weights represent the soft assignment of the stroke to each of the ACMs and depend on the distance between the stroke points and the ACM units, on gradient direction information and also on the stroke size. Both the weights and the ACMs energy minimization are computed using the generalized expectation-maximization (EM) algorithm. Initialization of the ACM's is fully automatic. Experimental results show the effectiveness of the proposed technique.

Autonomous detection of pulmonary nodules on CT images with a neural network-based fuzzy system

In this paper, a novel extension of neural network-based fuzzy model has been proposed to detect lung nodules. The proposed model can automatically identify a set of appropriate fuzzy inference rules, and refine the membership functions through the steepest gradient descent-learning algorithm. Twenty-nine clinical cases involving 583 thick section CT images were tested in this study. Receiver operating characteristic (ROC) analysis was used to evaluate the proposed autonomous pulmonary nodules detection system and yielded an area under the ROC curve of Azs=0.963. The overall detection sensitivity of the proposed method was 89.3% (with p-value less than 0.001), and the false positive was as low as 0.2 per image. This result demonstrates that the proposed neural network-based fuzzy system resolves the most suitable fuzzy rules, improves the detection rate, and reduces false positives compared to other approaches. The proposed system is fully automated with fast processing speed. The studies have shown a high potential for implementation of this system in clinical practice as a CAD tool.

Smoothing lung segmentation surfaces in three-dimensional X-ray CT images using anatomic guidance

RATIONALE AND OBJECTIVES

Automatic lung segmentation in volumetric computed tomography (CT) images has been extensively investigated, and several methods have been proposed. Most methods distinguish the lung parenchyma from the surrounding anatomy based on the difference in CT attenuation values. This leads to an irregular and inconsistent lung boundary for the regions near the mediastinum, which can cause inconsistent boundaries both across subjects and within subjects scanned at different intervals of time. Processes like lung image registration and lung atlas construction can be affected by such inconsistencies. Therefore there is a need for a more consistent lung surface near the mediastinum.

MATERIALS AND METHODS

This paper presents a fully automatic method for the three-dimensional smoothing of the lung boundary using information from the segmented human airway tree. First, using the segmented airway tree, we define a bounding box around the mediastinum for each lung, within which all operations are performed. We then define all generations of the airway tree distal to the right and left main stem bronchi to be part of the respective lungs and exclude all other segments. Finally, we perform a fast morphological closing with an ellipsoidal kernel to smooth the surface of the lung.

RESULTS

This method has been tested by processing the segmented lungs from eight normal datasets. The mean value of the magnitude of curvedness of the lung contours, averaged over all the datasets, is 0.13 postsmoothing, compared with 4.91 before smoothing. The accuracy of the lung contours after smoothing is assessed by comparing the automatic results to manually traced smooth lung borders by a human analyst. Averaged over all volumes, the root mean square difference between human and computer borders is 0.8691 mm after smoothing, compared with 1.3012 mm before smoothing. The mean similarity index, which is an area overlap measure based on the kappa statistic, is 0.9958 (SD 0.0032), after smoothing.

CONCLUSIONS

We have described a novel scheme for smoothing the lung contour around the mediastinum. The method is based on using anatomic information from the segmented airway tree. The validation results show that there is good agreement between manual and computer results. Because there are no accepted criteria for defining the lung boundary near the mediastinum, we believe our method of defining the boundary based on the structure of the airway tree provides a good basis for three-dimensional smoothing.

Toward automated segmentation of the pathological lung in CT

Conventional methods of lung segmentation rely on a large gray value contrast between lung fields and surrounding tissues. These methods fail on scans with lungs that contain dense pathologies, and such scans occur frequently in clinical practice. We propose a segmentation-by-registration scheme in which a scan with normal lungs is elastically registered to a scan containing pathology. When the resulting transformation is applied to a mask of the normal lungs, a segmentation is found for the pathological lungs. As a mask of the normal lungs, a probabilistic segmentation built up out of the segmentations of 15 registered normal scans is used. To refine the segmentation, voxel classification is applied to a certain volume around the borders of the transformed probabilistic mask. Performance of this scheme is compared to that of three other algorithms: a conventional, a user-interactive and a voxel classification method. The algorithms are tested on 10 three-dimensional thin-slice computed tomography volumes containing high-density pathology. The resulting segmentations are evaluated by comparing them to manual segmentations in terms of volumetric overlap and border positioning measures. The conventional and user-interactive methods that start off with thresholding techniques fail to segment the pathologies and are outperformed by both voxel classification and the refined segmentation-by-registration. The refined registration scheme enjoys the additional benefit that it does not require pathological (hand-segmented) training data.

Computer-aided lung nodule detection in CT: results of large-scale observer test

RATIONALE AND OBJECTIVES

The objective is to study the incremental effects of using a computer-aided lung nodule detection (CAD) system on the performance of a large pool of observers.

MATERIALS AND METHODS

A set of eight thin-section computed tomographic data sets with limited longitudinal coverage, containing a total of 22 lung nodules, was analyzed by using the automated nodule detection system. When applied to all eight cases, the CAD system alone achieved a detection rate of 86.4%, with 2.64 false-positive results per case. This study included 202 observers at a national radiology meeting: 39 thoracic radiologists, 95 non-thoracic radiologists, and 68 non-radiologists. Each participant read from one to eight cases in random order, first without and then with CAD system output available. Observer performance in nodule detection was measured before and after CAD was made available. Differences in performance of groups of observers before and after CAD were tabulated by mean, median, and SD in detection rate and number of false-positive results and tested by using nonparametric methods.

RESULTS

In an analysis involving only the first randomly selected case read by all 202 participants, there were statistically significant increases in nodule detection rates and numbers of false-positive results for all types of observers. There was a significant difference in detection rates between radiologists and non-radiologists before CAD, but after CAD, there was no significant difference in detection rates between these observer types. In a second analysis involving 13 participants who read all eight cases, mean detection rates were 64.0% before CAD and 81.9% after CAD. Mean numbers of false-positive results were 0.144 per case before CAD and 0.173 after CAD.

CONCLUSION

In a large observer study, use of a CAD system for nodule detection resulted in an incremental increase in detection rate, but also led to an increase in number of false-positive results. Also, CAD appears to be an equalizer of detection rates between observers of different levels of experience.

Database Design and Implementation for Quantitative Image Analysis Research

Quantitative image analysis (QIA) goes beyond subjective visual assessment to provide computer measurements of the image content, typically following image segmentation to identify anatomical regions of interest (ROIs). Commercially available picture archiving and communication systems focus on storage of image data. They are not well suited to efficient storage and mining of new types of quantitative data. In this paper, we present a system that integrates image segmentation, quantitation, and characterization with database and data mining facilities. The paper includes generic process and data models for QIA in medicine and describes their practical use. The data model is based upon the Digital Imaging and Communications in Medicine (DICOM) data hierarchy, which is augmented with tables to store segmentation results (ROIs) and quantitative data from multiple experiments. Data mining for statistical analysis of the quantitative data is described along with example queries. The database is implemented in PostgreSQL on a UNIX server. Database requirements and capabilities are illustrated through two quantitative imaging experiments related to lung cancer screening and assessment of emphysema lung disease. The system can manage the large amounts of quantitative data necessary for research, development, and deployment of computer-aided diagnosis tools.

Comparison of treatment response classifications between unidimensional, bidimensional, and volumetric measurements of metastatic lung lesions on chest computed tomography

RATIONALE AND OBJECTIVES

To study the agreement in treatment response classifications between unidimensional (1D), bidimensional (2D), and volumetric (3D) methods of measuring metastatic lung nodules on chest computed tomography (CT).

MATERIALS AND METHODS

Chest CT scans of 15 patients undergoing treatment for metastatic colorectal, renal cell, or breast carcinoma to the lungs were analyzed. CT images were acquired with 3 mm collimation and contiguous reconstruction. Two or three lung lesions were selected for each patient. Lesions were analyzed at baseline and two follow-up intervals of 1-4 months. 1D and 2D measurements were made with electronic calipers, while nodule volume was measured using a semiautomated segmentation system. Following the World Health Organization and RECIST (Response Evaluation Criteria in Solid Tumors) criteria, patients were categorized into four treatment response classifications. Volumetric criteria were used to classify response based on 3D measurements.

RESULTS

Thirty-two lesions from 15 patients were analyzed. Because each patient had a baseline and two follow-up scans, this yielded 30 response classifications for each measurement technique. The 1D, 2D, and 3D measurements were concordant in 21 of 30 classifications. The 1D and 3D measurements were concordant in 29 of 30 classifications, while the 2D and 3D measurements were concordant in 23 of 30 classifications. Level of agreement among the three methods was measured using a kappa statistic (K). For 1D compared with 3D, K = 0.739 +/- 0.345 (visits 1, 2) and 0.273 +/- 0.323 (visits 2, 3). For 2D compared with 3D, K = 0.655 +/- 0.325 (visits 1, 2) and 0.200 +/- 0.208 (visits 2, 3). Agreement among the methods for round and ovoid nodules was also fair to poor.

CONCLUSION

The three methods of tumor measurement show fair to poor agreement in treatment response classification. These findings have negative implications for the accuracy in which patients are classified under the World Health Organization or RECIST criteria and managed under cancer treatment protocols.

Segmentation of nodules on chest computed tomography for growth assessment

Several segmentation methods to evaluate growth of small isolated pulmonary nodules on chest computed tomography (CT) are presented. The segmentation methods are based on adaptively thresholding attenuation levels and use measures of nodule shape. The segmentation methods were first tested on a realistic chest phantom to evaluate their performance with respect to specific nodule characteristics. The segmentation methods were also tested on sequential CT scans of patients. The methods' estimation of nodule growth were compared to the volume change calculated by a chest radiologist. The best method segmented nodules on average 43% smaller or larger than the actual nodule when errors were computed across all nodule variations on the phantom. Some methods achieved smaller errors when examined with respect to certain nodule properties. In particular, on the phantom individual methods segmented solid nodules to within 23% of their actual size and nodules with 60.7 mm3 volumes to within 14%. On the clinical data, none of the methods examined showed a statistically significant difference in growth estimation from the radiologist.

Automated lung segmentation for thoracic CT impact on computer-aided diagnosis

RATIONALE AND OBJECTIVES

Automated lung segmentation in thoracic computed tomography scans is essential for the development of computer-aided diagnostic (CAD) methods. A core segmentation method may be developed for general application; however, modifications may be required for specific clinical tasks.

MATERIALS AND METHODS

An automated lung segmentation method has been applied (1) as preprocessing for automated lung nodule detection and (2) as the foundation for computer-assisted measurements of pleural mesothelioma tumor thickness. The core method uses gray-level thresholding to segment the lungs within each computed tomography section. The segmentation is revised through separation of right and left lungs along the anterior junction line, elimination of the trachea and main bronchi from the lung segmentation regions, and suppression of the diaphragm. Segmentation modifications required for nodule detection include a rolling ball algorithm to include juxtapleural nodules and morphologic erosion to eliminate partial volume pixels at the boundary of the segmentation regions.

RESULTS

For automated lung nodule detection, 4 of 82 actual nodules (4.9%) were excluded from the lung segmentation regions when the core segmentation method was modified compared with 14 nodules (17.1%) excluded without modifications. The computer-assisted quantification of mesothelioma method achieved a correlation coefficient of 0.990 with 134 manual measurements when the core segmentation method was used alone; correlation was reduced to 0.977 when the segmentation modifications, as adapted for the lung nodule detection task, were applied to the mesothelioma measurement task.

CONCLUSION

Different CAD applications impose different requirements on the automated lung segmentation process. The specific approach to lung segmentation must be adapted to the particular CAD task.

Quantitative evaluation of a pulmonary contour segmentation algorithm in X-ray computed tomography images

RATIONALE AND OBJECTIVES

Pulmonary contour extraction from thoracic x-ray computed tomography images is a mandatory preprocessing step in many automated or semiautomated analysis tasks. This study was conducted to quantitatively assess the performance of a method for pulmonary contour extraction and region identification.

MATERIALS AND METHODS

The automatically extracted contours were statistically compared with manually drawn pulmonary contours detected by six radiologists on a set of 30 images. Exploratory data analysis, nonparametric statistical tests, and multivariate analysis were used, on the data obtained using several figures of merit, to perform a study of the interobserver variability among the six radiologists and the contour extraction method. The intraobserver variability of two human observers was also studied.

RESULTS

In addition to a strong consistency among all of the quality indexes used, a wider interobserver variability was found among the radiologists than the variability of the contour extraction method when compared with each radiologist. The extraction method exhibits a similar behavior (as a pulmonary contour detector), to the six radiologists, for the used image set.

CONCLUSION

As an overall result of the application of this evaluation methodology, the consistency and accuracy of the contour extraction method was confirmed to be adequate for most of the quantitative requirements of radiologists. This evaluation methodology could be applied to other scenarios.

Multiparametric ISODATA analysis of embolic stroke and rt-PA intervention in rat

To increase the sensitivity of MRI parameters to detect tissue damage of ischemic stroke, an unsupervised analysis method, Iterative Self-Organizing Data Analysis Technique Algorithm (ISODATA), was applied to analyze the temporal evolution of ischemic damage in a focal embolic cerebral ischemia model in rat with and without recombinant tissue plasminogen activator (rt-PA) treatment. Male Wistar rats subjected to embolic stroke were investigated using a 7-T MRI system. Rats were randomized into control (n=9) and treated (n=9) groups. The treated rats received rt-PA via a femoral vein at 4 h after onset of embolic ischemia. ISODATA analysis employed parametric maps or weighted images (T1, T2, and diffusion). ISODATA results with parametric maps are superior to ISODATA with weighted images, and both of them were highly correlated with the infarction size measured from the corresponding histological section. At 24 h after embolic stroke, the average map ISODATA lesion sizes were 37.7+/-7.0 and 39.2+/-5.6 mm2 for the treated and the control group, respectively. Average histological infarction areas were 37.9+/-7.4 mm2 for treated rats and 39.4+/-6.1 mm2 for controls. The R2 values of the linear correlation between map ISODATA and histological data were 0.98 and 0.96 for treated and control rats, respectively. Both histological and map ISODATA data suggest that there is no significant difference in infarction area between non-treated and rt-PA-treated rats when treatment was administered 4 h after the onset of embolic stroke. The ISODATA lesion size analysis was also sensitive to changes of lesion size during acute and subacute stages of stroke. Our data demonstrate that the multiparameter map ISODATA approach provides a more sensitive quantitation of the ischemic lesion at all time points than image ISODATA and single MRI parametric analysis using T1, T2 or ADCw.

Emphysema: effect of reconstruction algorithm on CT imaging measures

In the current study, the effects of reconstruction algorithms on quantitative measures derived from computed tomographic (CT) lung images were assessed in patients with emphysema. CT image data sets were reconstructed with a standard algorithm and alternative algorithm(s) for 42 subjects. Algorithms were grouped as overenhancing, sharp, standard, or smooth. Density mask and volume measurements from the alternative algorithm data sets were compared with standard algorithm data sets. The overenhancing category yielded an average shift of 9.4% (ie, a shift in average score from 35.5% to 44.9%); the sharp category, a shift of 2.4%; and the smooth category, a shift of -1.0%. Differences in total lung volume measurements were less than 1%. In conclusion, the CT reconstruction algorithm may strongly affect density mask results, especially for certain reconstruction algorithms.

Segmentation of magnetic resonance images using a combination of neural networks and active contour models

Segmentation of medical images is very important for clinical research and diagnosis, leading to a requirement for robust automatic methods. This paper reports on the combined use of a neural network (a multilayer perceptron, MLP) and active contour model ('snake') to segment structures in magnetic resonance (MR) images. The perceptron is trained to produce a binary classification of each pixel as either a boundary or a non-boundary point. Subsequently, the resulting binary (edge-point) image forms the external energy function for a snake, used to link the candidate boundary points into a continuous, closed contour. We report here on the segmentation of the lungs from multiple MR slices of the torso; lung-specific constraints have been avoided to keep the technique as general as possible. In initial investigations, the inputs to the MLP were limited to normalised intensity values of the pixels from an (7 x 7) window scanned across the image. The use of spatial coordinates as additional inputs to the MLP is then shown to provide an improvement in segmentation performance as quantified using the effectiveness measure (a weighted product of precision and recall). Training sets were first developed using a lengthy iterative process. Thereafter, a novel cost function based on effectiveness is proposed for training that allows us to achieve dramatic improvements in segmentation performance, as well as faster, non-iterative selection of training examples. The classifications produced using this cost function were sufficiently good that the binary image produced by the MLP could be post-processed using an active contour model to provide an accurate segmentation of the lungs from the multiple slices in almost all cases, including unseen slices and subjects.

Automated lung segmentation in X-ray computed tomography: development and evaluation of a heuristic threshold-based scheme

RATIONALE AND OBJECTIVES

To develop and evaluate a reliable, fully-automated lung segmentation scheme for application in X-ray computed tomography.

MATERIALS AND METHODS

The automated scheme was heuristically developed using a slice-based, pixel-value threshold and two sets of classification rules. Features used in the rules include size, circularity, and location. The segmentation scheme operates slice-by-slice and performs three key operations: (1) image preprocessing to remove background pixels, (2) computation and application of a pixel-value threshold to identify lung tissue, and (3) refinement of the initial segmented regions to prune incorrectly detected airways and separate fused right and left lungs.

RESULTS

The performance of the automated segmentation scheme was evaluated using 101 computed tomography cases (91 thick slice, 10 thin slice scans). The 91 thick cases were pre- and post-surgery from 50 patients and were not independent. The automated scheme successfully segmented 94.0% of the 2,969 thick slice images and 97.6% of the 1,161 thin slice images. The mean difference of the total lung volumes calculated by the automated scheme and functional residual capacity plus 60% inspiratory capacity was -24.7 +/- 508.1 mL. The mean differences of the total lung volumes calculated by the automated scheme and an established, commonly used semi-automated scheme were 95.2 +/- 52.5 mL and -27.7 +/- 66.9 mL for the thick and thin slice cases, respectively.

CONCLUSION

This simple, fully-automated lung segmentation scheme provides an objective tool to facilitate lung segmentation from computed tomography scans.

Lung nodule detection and characterization with multislice CT

The ability to identify and characterize pulmonary nodules has been dramatically increased by the introduction of multislice CT (MSCT) technology. Using high-resolution sections, MSCT allows considerable improvement in assessing nodule morphology, enhancement patterns, and growth. MSCT also has facilitated the development and potential of clinical application of computer-assisted diagnosis.

Establishing a normative atlas of the human lung: intersubject warping and registration of volumetric CT images

RATIONALE AND OBJECTIVES

To establish the range of normal values for quantitative CT-based measures of lung structure and function, the authors developed a method for matching pulmonary structures across individuals and creating a normative human lung atlas.

MATERIALS AND METHODS

A computerized human lung atlas was synthesized from computed tomographic (CT) images from six subjects by means of three-dimensional image registration. The authors identified a set of reproducible feature points for each CT image and used these points to establish correspondences across subjects, used a landmark- and intensity-based consistent image registration algorithm to register a template image volume from the population to the rest of the pulmonary CT volumes in the population, averaged these transformations, and constructed an atlas by deforming the template with the average transformation.

RESULTS

The effectiveness of the authors' method was evaluated and visualized by means of both gray-level and segmented CT images. The method reduced the average landmark registration error from 10.5 mm to 0.4 mm and the average relative volume overlap error from 0.7 to 0.11 for the six data sets studied.

CONCLUSION

The method, and the computerized human lung atlas constructed and visualized by the authors with this method, provides a basis for establishing regional ranges of normative values for structural and functional measures of the human lung.

Lung micronodules: automated method for detection at thin-section CT--initial experience

An automated system was developed for detecting lung micronodules on thin-section computed tomographic images and was applied to data from 15 subjects with 77 lung nodules. The automated system, without user interaction, achieved a sensitivity of 100% for nodules (>3 mm in diameter) and 70% for micronodules (<or=3 mm). With the same images, a radiologist detected nodules and micronodules with sensitivities of 91% and 51%, respectively, without system input. With assistance from the automated system, these sensitivities increased to 95% and 74%, respectively. Preliminary results indicate that the automated system considerably improved the radiologist's performance in micronodule detection.

Lung nodule detection on thoracic computed tomography images: preliminary evaluation of a computer-aided diagnosis system

We are developing a computer-aided diagnosis (CAD) system for lung nodule detection on thoracic helical computed tomography (CT) images. In the first stage of this CAD system, lung regions are identified by a k-means clustering technique. Each lung slice is classified as belonging to the upper, middle, or the lower part of the lung volume. Within each lung region, structures are segmented again using weighted k-means clustering. These structures may include true lung nodules and normal structures consisting mainly of blood vessels. Rule-based classifiers are designed to distinguish nodules and normal structures using 2D and 3D features. After rule-based classification, linear discriminant analysis (LDA) is used to further reduce the number of false positive (FP) objects. We performed a preliminary study using 1454 CT slices from 34 patients with 63 lung nodules. When only LDA classification was applied to the segmented objects, the sensitivity was 84% (53/63) with 5.48 (7961/1454) FP objects per slice. When rule-based classification was used before LDA, the free response receiver operating characteristic (FROC) curve improved over the entire sensitivity and specificity ranges of interest. In particular, the FP rate decreased to 1.74 (2530/1454) objects per slice at the same sensitivity. Thus, compared to FP reduction with LDA alone, the inclusion of rule-based classification lead to an improvement in detection accuracy for the CAD system. These preliminary results demonstrate the feasibility of our approach to lung nodule detection and FP reduction on CT images.

Medical image segmentation with knowledge-guided robust active contours

Medical image segmentation techniques typically require some form of expert human supervision to provide accurate and consistent identification of anatomic structures of interest. A novel segmentation technique was developed that combines a knowledge-based segmentation system with a sophisticated active contour model. This approach exploits the guidance of a higher-level process to robustly perform the segmentation of various anatomic structures. The user need not provide initial contour placement, and the high-level process carries out the required parameter optimization automatically. Knowledge about the anatomic structures to be segmented is defined statistically in terms of probability density functions of parameters such as location, size, and image intensity (eg, computed tomographic [CT] attenuation value). Preliminary results suggest that the performance of the algorithm at chest and abdominal CT is comparable to that of more traditional segmentation techniques like region growing and morphologic operators. In some cases, the active contour-based technique may outperform standard segmentation methods due to its capacity to fully enforce the available a priori knowledge concerning the anatomic structure of interest. The active contour algorithm is particularly suitable for integration with high-level image understanding frameworks, providing a robust and easily controlled low-level segmentation tool. Further study is required to determine whether the proposed algorithm is indeed capable of providing consistently superior segmentation.

A pilot study of all-trans-retinoic acid for the treatment of human emphysema

Emphysema results from progressive destruction of alveolar septae and was considered irreversible until all-trans-retinoic acid (ATRA) was shown to reverse anatomic and physiologic signs of emphysema in a rat model. To evaluate the feasibility of ATRA as a clinical therapy, 20 patients with severe emphysema were enrolled into a randomized, double-blind, placebo-controlled pilot study. Participants included 16 male and 4 female former smokers, two with alpha(1)-antitrypsin deficiency. Patients were treated with either 3 mo of ATRA (50 mg/m(2)/d) or 3 mo of placebo, followed by a 3-mo crossover phase. Plasma drug levels were followed and outcome measures included serial pulmonary function tests, blood gases, lung compliance, computed tomography (CT) imaging, and quality of life questionnaires. In general, treatment was well tolerated and associated with only mild side effects including skin changes, transient headache, hyperlipidemia, transaminites, and musculoskeletal pains. Plasma drug levels varied considerably between subjects and decreased significantly over time in 35% of the participants. Physiologic and CT measurements did not change appreciably in response to therapy. We conclude that ATRA is well tolerated in patients with emphysema, and trials evaluating higher doses, longer treatment, or different dosing schedules are feasible.

Patient-specific models for lung nodule detection and surveillance in CT images

The purpose of this work is to develop patient-specific models for automatically detecting lung nodules in computed tomography (CT) images. It is motivated by significant developments in CT scanner technology and the burden that lung cancer screening and surveillance imposes on radiologists. We propose a new method that uses a patient's baseline image data to assist in the segmentation of subsequent images so that changes in size and/or shape of nodules can be measured automatically. The system uses a generic, a priori model to detect candidate nodules on the baseline scan of a previously unseen patient. A user then confirms or rejects nodule candidates to establish baseline results. For analysis of follow-up scans of that particular patient, a patient-specific model is derived from these baseline results. This model describes expected features (location, volume and shape) of previously segmented nodules so that the system can relocalize them automatically on follow-up. On the baseline scans of 17 subjects, a radiologist identified a total of 36 nodules, of which 31 (86%) were detected automatically by the system with an average of 11 false positives (FPs) per case. In follow-up scans 27 of the 31 nodules were still present and, using patient-specific models, 22 (81%) were correctly relocalized by the system. The system automatically detected 16 out of a possible 20 (80%) of new nodules on follow-up scans with ten FPs per case.