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.     

A three-object model for the similarity searches of chest CT images

We propose in this paper a three-object model specifically for the archiving and retrieval of chest CT images. To calculate parameters for the model, each chest CT image needs to be processed to segment the three main objects and then the features be extracted to describe the objects' properties and relationships. In the image segmentation part, we applied the knowledge of the modality on chest CT images and modified the traditional watershed image segmentation algorithm including a four-step merging algorithm specifically for chest CT images. After segmentation, the mediastinum and two lung lobes are identified. The mediastinum object is mainly described by shape-related features while the two lung lobes are described mainly by texture features. A three-object model was exploited to describe the object features and the spatial relationship among objects.

To test the capability of the three-object model to the similarity searches of chest CT images, we developed a CBIR system in which three distinct query modes were provided. They are ‘searching by ARGs’, ‘searching by shape features of mediastinum’, and ‘searching by texture features of lung lobes’. The experimental results show that the three-object model demonstrates impressive power in the similarity searching of chest CT images. Among the three searching modes, the ‘searching by shape features of mediastinum’ and ‘searching by texture features of lung lobes’ modes provide user choices to search for images with high similarities in specific objects rather than in the whole images. The precision rate of either query mode is high, with an average of around 80% out of the first 30 result images are justified as similar, which is impressive in a fully automatic image query system using content features. Nevertheless, the two query modes that concentrate on distinct object features show slightly better capability in searching for similar images than the ‘searching by ARGs’ mode.     

Development of computer-aided diagnostic system for detection of lung nodules in three-dimensional computed tomography images

We have developed an automated computerized method for the detection of lung nodules in three-dimensional (3D) computed tomography (CT) images obtained by helical CT. In this scheme, a lung segmentation technique for the determination of the nodule search area is performed based on a gray-level thresholding technique. To enhance lung nodules, we employed the 3D cross-correlation method by using a 3D Gaussian template with zero-surrounding as a model of lung nodule. False positives are then eliminated by using a rule-base with 53 features. For further reduction of false positives, we performed linear discriminant analysis using these 53 features. The average number of false positives was 6.7 per case at a percent sensitivity of 85.0%. This computerized scheme will be useful to radiologists by providing a "second opinion" in case of possible early lung cancer.     

胸部撞击伤复合脂多糖致急性肺损伤大鼠模型的建立

目的建立SD（SpragueDawley）大鼠经胸部撞击复合内毒素所致急性肺损伤（ALI）模型,为后续Au的救治和发病机制研究提供支撑。方法采用BMI-Ⅲ型生物撞击机,对SD大鼠行右侧胸部准静态正面撞击,比较动物受撞击后的呼吸及伤后24h肺部大体伤情。确认25kPa为合适的驱动压力后,于撞击后即刻通过尾静脉分别注射10、15、20mg／kg3种剂量脂多糖（LPS）,观察伤后动物的死亡率及肺损伤病理情况。结果250kPa所致肺损伤伤情中度,动物无死亡。撞击后即刻复合LPS的3种剂量中,20rag／kg可致动物48h死亡率为80％,10rag／kg则致动物48h死亡率仅20％,15mg／kg致动物48h死亡率为70％。250kPa＋15mg／kg条件下的肺组织病理检测可见撞击区旁红细胞、炎细胞聚集,肺泡腔及肺泡间隔出现粉红色液体,肺泡间隔增厚等ALl典型病变。结论以250kPa驱动压力行右侧胸部撞击复合15mg／kg的LPS静脉注射,既保证伤情达到一定严重程度,又保证后续能得到有效的治疗观察结果,为最佳致伤条件。所建立的大鼠ALl动物模型操作简便,致伤参数可控,且重复性好,病变典型,可作为研究急性肺损伤较理想的实马龠动物樽型．     

Improved Method for Automatic Identification of Lung Regions on Chest Radiographs

RATIONALE AND OBJECTIVES: The authors performed this study to evaluate an algorithm developed to help identify lungs on chest radiographs. MATERIALS AND METHODS: Forty clinical posteroanterior chest radiographs obtained in adult patients were digitized to 12-bit gray-scale resolution. In the proposed algorithm, the authors simplified the current approach of edge detection with derivatives by using only the first derivative of the horizontal and/or vertical image profiles. In addition to the derivative method, pattern classification and image feature analysis were used to determine the region of interest and lung boundaries. Instead of using the traditional curve-fitting method to delineate the lung, the authors applied an iterative contour-smoothing algorithm to each of the four detected boundary segments (costal, mediastinal, lung apex, and hemidiaphragm edges) to form a smooth lung boundary. RESULTS: The algorithm had an average accuracy of 96.0% for the right lung and 95.2% for the left lung and was especially useful in the delineation of hemidiaphragm edges. In addition, it took about 0.775 second per image to identify the lung boundaries, which is much faster than that of other algorithms noted in the literature. CONCLUSION: The computer-generated segmentation results can be used directly in the detection and compensation of rib structures and in lung nodule detection.     

基于薄层CT的影像组学和形态学特征联合模型在预测磨玻璃样肺腺癌浸润程度的价值

目的 探讨基于薄层CT的影像组学和形态学特征联合模型在预测磨玻璃样肺腺癌中的原位癌(AIS)、微浸润腺癌(MIA)、浸润性腺癌(IAC)的价值.方法 回顾性分析2018年6月至2021月3月经病理证实的327例肺腺癌患者(335个病灶)术前肺部CT检查图像,随机选取201个为训练集,134个为测试集.肺腺癌参照病理金标...     

Automatic segmentation and recognition of lungs and lesion from CT scans of thorax

In this study, a fully automated texture-based segmentation and recognition system for lesion and lungs from CT of thorax is presented. For the segmentation part, we have extracted texture features by Gabor filtering the images, and, then combined these features to segment the target volume by using Fuzzy C Means (FCM) clustering. Since clustering is sensitive to initialization of cluster prototypes, optimal initialization of the cluster prototypes was done by using a Genetic Algorithm.For the recognition stage, we have used cortex like mechanism for extracting statistical features in addition to shape-based features. The segmented regions showed a high degree of imbalance between positive and negative samples, so we employed over and under sampling for balancing the data. Finally, the balanced and normalized data was subjected to Support Vector Machine (SimpleSVM) for training and testing.Results reveal an accuracy of delineation to be 94.06%, 94.32% and 89.04% for left lung, right lung and lesion, respectively. Average sensitivity of the SVM classifier was seen to be 89.48%.     

123I-MIBG pulmonary removal: a biochemical marker of minimal lung endothelial cell lesions.

Circulating biogenic amines, such as serotonin or norepinephrine, are cleared by the lung and can be considered as indicators of pulmonary endothelial integrity. An iodinated norepinephrine analogue, metaiodobenzylguanidine (MIBG), is extracted in the lung by an active, saturable transport system. In order to investigate the use of MIBG lung extraction to detect minimal lung endothelial cell lesions, an experimental model of initial pulmonary vascular lesions was developed in rats using repeated daily intraperitoneal injection of bleomycin (BLM: 2 U/100 g body weight) over a 5-day period. At this time, a significant decrease of serum angiotensin-converting enzyme activity, an index of endothelial cell metabolism, was observed (214 +/- 11 U/ml in controls as compared with 182 +/- 11 U/ml in BLM-treated rats, P less than 0.05); electron microscopy showed the presence of typical but minimal endothelial cell lesions (blebbing). MIBG extraction (%EMIBG) was measured using the isolated perfused lung model after a 10-min steady-state period and 2 min of perfusion with 123I-MIBG (0.1 microCi/ml) and 125I-labelled human serum albumin (HSA) (0.2 microCi/ml). Intraperitoneal administration of BLM to rats for 5 days resulted in a significant decrease in pulmonary extraction of MIBG. %EMIBG declined from a control value of 24.7% +/- 1.1% in controls to 19.3 +/- 1.6% in BLM-treated rats (n = 27, P less than 0.05; -21.2%). HSA lung extraction, used as an estimate of nonspecific residual lung activity, was not different in the lungs of BLM-treated rats as compared with controls.(ABSTRACT TRUNCATED AT 250 WORDS)     

经胸超声心动图对纵隔占位性病变的诊断价值

目的探讨纵隔占位性病变的超声特点,以提高早期诊断率。方法回顾性分析我院经胸超声心动图(TTE)检出并经病理或临床结合CT证实的纵隔占位性病变7例,扫查心脏左室长轴、胸骨上窝、剑下切面,重点观察心包周围、升主动动脉前方区域、降主动脉两侧和心脏后方。结果7例中,前纵隔占位4例;3例为囊性占位,分别为皮样囊肿、胸腺囊肿和肺癌并包裹性纵隔胸膜积液;胸腺囊肿及皮样囊肿均边界清,内部透声好。肺癌并包裹性胸腔积液形态欠规则,后壁可见松软低回声物质积聚。实性占位1例,为胸内甲状腺肿,边界清,呈实性不均质回声。后纵隔占位3例,其中神经纤维瘤1例,肝癌、肺癌并后纵隔转移各1例。神经纤维瘤呈中等回声的实质性团块,边界清晰,其内回声欠均匀;肝癌、肺癌并后纵隔转移边缘不规则,呈实性不均质团块回声。结论根据TTE检查切面的特点,结合纵隔占位的图像特征,提高TTE医师对纵隔内占位性病变的扫查意识,有助于纵隔内占位性病变的早期诊断和治疗方案的制定。     

Advanced emphysema: preoperative chest radiographic findings as predictors of outcome following lung volume reduction surgery.

PURPOSE: To determine whether preoperative chest radiographic findings alone can reliably predict which patients will achieve the best functional outcome of lung volume reduction surgery. MATERIALS AND METHODS: The preoperative chest radiographs obtained in 57 patients who had undergone lung volume reduction surgery were retrospectively scored by five blinded readers for severity and distribution of emphysema, evidence of lung compression, disease heterogeneity, and other features. Comparisons were made with the 3-6-month postoperative functional outcome for each patient. RESULTS: High disease heterogeneity (score > 2) and unequivocal lung compression (score 1) both were 100% predictive of a favorable outcome (FEV1 increase, > or = 30%). Low heterogeneity (score < 1) was 94% predictive of an unfavorable outcome (FEV1 increase < 30%), as was a lack of lung compression, which was 92% predictive of an unfavorable outcome. These two features also correlated with an improved 6-minute walk test result, although this correlation was weaker. CONCLUSION: Chest radiography alone may be sufficient for initial screening. High disease heterogeneity and lung compression on chest radiographs are highly predictive of a favorable functional outcome.     

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.     

Image annotation for conveying automated lung nodule detection results to radiologists

RATIONALE AND OBJECTIVES: The author investigated the ability of automated techniques to convey the results of an automated lung nodule detection method for human visualization. MATERIALS AND METHODS: Automated nodule detection begins with gray-level thresholding techniques to create a segmented lung volume within which nodule candidates are identified. Morphologic and gray-level features are computed for each candidate. To distinguish between candidates that represent actual nodules and those that represent non-nodules, a rule-based scheme is combined with linear discriminant analysis. For output visualization, final detection results are represented as circles around computer-detected structures in a single section in which each structure appears. Consequently, an inappropriate choice of section could result in an actual nodule detected by the computer but not properly indicated to the radiologist, thus reducing the potential positive impact of that detection on the radiologist's decision-making process. RESULTS: The automated nodule detection method achieved 71% sensitivity with 0.5 false positives per section on 38 CT scans; however, when these results were converted to annotations on the images output for human visualization, only 91% of the computer-detected true-positive nodules received annotations that encompassed a portion of the actual nodule. Thus, the "effective sensitivity" of the automated detection method was reduced. CONCLUSION: The "effective sensitivity" of an automated lung nodule detection system considers the eventual human interaction with system output. Differences between reported computer sensitivity and "effective sensitivity" may be reduced through proper consideration of the assessment of "truth," of the manner in which computer results are scored, and of the complete segmentation of candidates for automated nodule detection.     

Pulmonary nodule characterization: a comparison of conventional with quantitative and visual semi-quantitative analyses using contrast enhancement maps

PURPOSE: To determine whether conventional nodule densitometry or analysis based on contrast enhancement maps of indeterminate lung nodules imaged with contrast-enhanced CT can distinguish benign from malignant lung nodules. MATERIALS AND METHOD: Thin section, contrast-enhanced CT (baseline, and post-contrast series acquired at 45, 90,180, and 360 s) was performed on 29 patients with indeterminate lung nodules (14 benign, 15 malignant). A thoracic radiologist identified the boundary of each nodule using semi-automated contouring to form a 3D region-of-interest (ROI) on each image series. The post-contrast series having the maximum mean enhancement was then volumetrically registered to the baseline series. The two series were subtracted volumetrically and the subtracted voxels were quantized into seven color-coded bins, forming a contrast enhancement map (CEM). Conventional nodule densitometry was performed to obtain the maximum difference in mean enhancement values for each nodule from a circular ROI. Three thoracic radiologists performed visual semi-quantitative analysis of each nodule, scoring each map for: (a) magnitude and (b) heterogeneity of enhancement throughout the entire volume of the nodule on a five-point scale. Receiver operator characteristic (ROC) analysis was conducted on these features to evaluate their diagnostic efficacy. Finally, 14 quantitative texture features were calculated for each map. A statistical analysis was performed to combine the 14 texture features to a single factor. ROC analysis of the derived aggregate factor was done as an indicator of malignancy. All features were analyzed for differences between benign and malignant nodules. RESULTS: Using 15 HU as a threshold, 93% (14/15) of malignant and 79% (11/14) of benign nodules demonstrated enhancement. The ROC curve when higher values of enhancement indicate malignancy was generated and area under the curve (AUC) was 0.76. The visually scored magnitude of enhancement was found to be less effective in distinguishing malignant from benign lesions, with an average AUC of 0.62. The visually scored pattern of enhancement was found to be more effective with an average AUC of 0.79. From the statistical analysis performed to combine the texture features to a single factor, the area under the ROC curve was 0.84. CONCLUSION: The present study suggests that visual semi-quantitative and quantitative characterization of contrast enhancement patterns may potentially enhance the discrimination between benign and malignant nodules. Further studies and correlation with pathologic material will be important to better understand the potential interplay between CT enhancement features, host stromal elements, and neovascularity that may contribute to these patterns.     

The appearance of liquid surfaces and layers in routine radiographs

Image features in the radiograph produced by deformation of a liquid surface by surface tension and by the density gradient in a diffusion layer may present unexpected difficulty of interpretation. Such features have been analysed in model experiments, which have been reported earlier. The aim of the present investigation was to examine the occurrence and the clinical implications of corresponding phenomena in routine radiographs. In the human body liquid surfaces and diffusion layers can occur only in cavities, both normal and abnormal. A liquid surface tends to extend up a cavity wall to form a meniscoid or, if the cavity is small enough, a discoid. The liquid surface continues further up the wall as a liquid film. The shape of the meniscoid and the discoid varies with the shape and inclination of the wall. Most of the image features of interest are produced by rays that are tangential to a horizontal surface, a meniscoid, a discoid or a concave wall, any of which is visualized as an internal boundary with a light Mach line. When the wall is convex towards the cavity the meniscoid is saddle-shaped and an external boundary with a dark Mach line is produced. The horizontal part of a liquid surface can be touched only if it is at the same level as the focus of the roentgen tube. A liquid surface at any other level can be touched in its meniscoid only by rays that are not horizontal. It is reproduced as an internal boundary, slightly concave upwards; above this boundary the rest of the liquid surface is reproduced as a wedge field.(ABSTRACT TRUNCATED AT 250 WORDS)     

Comparison of a lung fitting algorithm with CT data for tangential fields in radiotherapy of the breast

A method of estimating the shape and position of the lung in tangential breast fields is presented for patients who have not been CT scanned. Using the Osiris system, the external contour is obtained optically, and an estimated lung structure superimposed on the transverse outlines based on the measured lung depth in the tangential fields and an analysis of the typical lung shapes obtained from CT images. The accuracy of this fit was determined by comparison with a set of 64 CT images imported into the Osiris system. Dose distributions were calculated by two treatment planning systems: ADAC Pinnacle and GE Target2. The computed dose distributions for 6 MV photons were compared against measured doses in a specialized breast phantom. For the worst case of lung fit compared with CT, the dosimetric error (based upon ADAC Pinnacle calculations) was 2.0% in the shadow of the lung. For the complete patient data set, the relative dose errors to these points were reduced from a mean value of 8.4% and standard deviation (SD)=1.8% (no lung correction) to a mean of 0.2% and SD=1.0% (lung correction using fitted lung). It was also found that for every 1 cm of lung path length the dose to the breast along that path length increased by approximately 1%. The results of these investigations indicated that the lung fit model was satisfactory for routine clinical use, so that good dosimetric results can be obtained using lung correction without the need for CT imaging.     

风湿性肺病的HRCT表现

目的:探讨风湿性肺部病变HRCT表现的一般特征。材料和方法:分析308例风湿病的肺HRCT图像,观察病变的分布特点、肺与支气管病变的形态特点。结果:风湿性肺病的病变分布主要为周边分布(63.6％)、偏后侧分布(57.8％)及膈上肺底分布(61.4％)。本组大多数肺部病变表现为磨玻璃影(68.9％)及线样结构(59.4％),部分病例可伴有网状改变或蜂窝影(32.2％)及界面征(34.1％)。风湿病大支气管及小支气管的受累率分别为1.9％和32.5％(χ~2=96.54,P<0.01)。马赛克灌注的检出率在吸气、呼气相扫描上分别为15.3％、25.6％(χ~2=10.35,P<0.01)。结论:风湿性肺病HRCT表现有一定的特征。病变分布主要位于周边、偏后侧及膈上区域;肺部病变表现主要为磨玻璃影、线样结构、网状改变或蜂窝影及界面征等;支气管病变主要累及小支气管。     

常见恶性肿瘤颅内转移的MRI征象分析

目的：探讨常见恶性肿瘤颅内转移的M RI特点。方法回顾性分析210例均接受M RI平扫及增强扫描,临床确诊为脑转移瘤且已明确原发肿瘤类型的患者资料,并作统计分析。结果脑转移瘤的数目、大小、强化形态、瘤周水肿均与原发肿瘤相关（ P ＜0．05）。MRI扫描中,肺癌、乳腺癌等常见恶性肿瘤颅内转移多表现为长T1、长 T2信号,结直肠癌转移灶中有20个（71．4％）呈混杂/短T2信号,DWI上245个（69．4％）转移灶边界清楚。45例（70．3％）肺腺癌、33例（55．0％）肺小细胞癌及15例（78．9％）乳腺癌颅内转移,M RI表现以多发转移灶居多;16例（59．3％）肺鳞癌以单发转移灶居多。72个（57．1％）肺小细胞癌的转移灶为结节状,22个（59．5％）肺鳞癌的转移灶为粟粒状;肺鳞癌有27个（72.9％）转移灶为环形强化;18个（48．6％）肺鳞癌转移灶的瘤周水肿为轻度,重度水肿中有21个（52．5％）转移灶为肺腺癌。结论不同恶性肿瘤颅内转移可有不同的M RI特征;M RI平扫、增强及DWI综合运用对脑转移瘤的诊断及鉴别有重要价值。     

CT-findings in ARDS

The CT features of 28 patients with ARDS are described. Diffuse lung consolidation, multifocal patchy involvement and lobar or segmental disease were observed. Large lung cysts as well as small cysts producing a "swiss-cheese" appearance of the parenchyma, were detected. These findings were not regularly appreciated on chest radiographs. The overall mortality of our 28 patients was 72.7% (22 out of 28). Patients with lung cysts showed a trend toward higher mortality (87.5% or 13 out of 16). Other unexpected findings were basilar lung abscesses and an empyema. In 15 out of 28 patients, CT scans provided additional information, not obvious on bedside chest radiographs and led to a change in management in five patients.     

Computer-aided classification of interstitial lung diseases via MDCT: 3D adaptive multiple feature method (3D AMFM)

RATIONALE AND OBJECTIVES: Computer-aided detection algorithms applied to multidetector row CT (MDCT) lung image data sets have the potential to significantly alter clinical practice through the early, quantitative detection of pulmonary pathology. In this project, we have further developed a computer-aided detection tool, the adaptive multiple feature method (AMFM), for the detection of interstitial lung diseases based on MDCT-generated volumetric data. MATERIALS AND METHODS: We performed MDCT (Siemens Sensation 16 or 64 120 kV, B50f convolution kernel, and 相似文献   

An edge-region force guided active shape approach for automatic lung field detection in chest radiographs

Automatic and accurate lung field segmentation is an essential step for developing an automated computer-aided diagnosis system for chest radiographs. Although active shape model (ASM) has been useful in many medical imaging applications, lung field segmentation remains a challenge due to the superimposed anatomical structures. We propose an automatic lung field segmentation technique to address the inadequacy of ASM in lung field extraction. Experimental results using both normal and abnormal chest radiographs show that the proposed technique provides better performance and can achieve 3-6% improvement on accuracy, sensitivity and specificity compared to traditional ASM techniques.