Reducing FPs in Nodule Detection Using Neural Networks Ensemble

We have developed a novel method called local contralateral subtraction for reduction of false positives reported by a computer-aided diagnosis (CAD) scheme for detection of lung nodules in chest radiographs. Our method is based on the removal of normal structures in the regions of interest (ROIs), based on symmetry between the left and right lungs. In our method, two ROIs were extracted, one from the position where a candidate of a nodule is located, and the other from the anatomically corresponding location in the opposite lung, which contains similar normal structures. We employed a wavelet-based multiresolution image registration method to match the two ROIs, and subtraction was performed. A signal- to-noise ratio (SNR) between a central region and the adjacent background region was calculated for quantification of the remaining structures in the subtracted ROI. The SNR was then used for distinction between nodules and false positives. In an analysis of 550 ROIs consisting of 51 nodules and 499 false positives reported as detected nodules by our current CAD scheme, we were able to eliminate 44% of false positives with loss of only one nodule with this new method. This result indicates that our new method is effective in reducing false positives due to normal anatomic structures, and thus can improve the performance of our CAD scheme for detection of pulmonary nodules in chest radiographs.

Computerized detection of pulmonary nodules in chest radiographs based on morphological features and wavelet snake model

Observer studies are frequently performed to test new modalities. Correct study design is important to generate reliable results. Two most frequently used observer study designs are the sequential and the independent reading design. We investigated the effect of different observer study designs on reader performance results and statistical power. The study included multiple assessments of chest radiographs (CXR) with bone suppression images (BSI) for the detection of lung nodules. In a fully crossed study design 8 observers assessed first radiographs without and with BSI sequentially. Secondly they scored radiographs independently having BSI available from the beginning. Five months later, the same readers scored the same cases again in an independent reading session, completing the three scorings for CXRs with BSI. Observer performance was compared using multi reader multi case (MRMC) receiver operating characteristics (ROC). To estimate reader variance, Dorfman, Berbaum, Metz (DBM) variance component estimates were calculated. No significant difference between the sequential and the independent reading sessions could be found (p=0.51; p=0.61). Both reading designs showed increased performance with BSI, with a significant increase for the sequential and the independent reading session after five months (p=0.002; p=0.007). Total observer variance between sequential and independent reading design remained the same. A strong increase of uncorrelated components was found in the independent reading sessions, masking the ability to demonstrate differences in observer performance across modalities. In conclusion, results of the sequential and the independent study design did not show a significant difference. The independent study design had less power compared to the sequential study design due to a strong increase of uncorrelated variance components.

A computer-aided diagnosis system for detection of lung nodules in chest radiographs with an evaluation on a public database

Computer-vision techniques based on wavelets are presented with applications to computer-aided diagnosis (CAD) in medical images. CAD is a method for assisting radiologists’ interpretation of medical images. CAD schemes act as a second reader and alert radiologists to suspicious lesions. Development of CAD schemes is an active area of research, motivated by the fact that radiologists tend to misdiagnose cancerous lesions in medical images. In the past several years, we have been extensively developing wavelet-based techniques for CAD schemes. Development of these techniques has led to a new CAD scheme and has yielded significant improvements in the performance of the existing CAD schemes. This chapter describes two wavelet-based, low-level computer vision techniques: simultaneous segmentation and registration, and matching pursuit with optimally weighted wavelet packets. We present their application to the computer-aided detection of lung nodules in chest radiographs and microcalcifications in mammograms.

Early detection of lung cancer is crucial in reducing mortality. Magnetic resonance imaging (MRI) may be a viable imaging technique for lung cancer detection. Numerous lung nodule detection methods have been studied for computed tomography (CT) images. However, to the best of our knowledge, no detection methods have been carried out for the MR images. In this paper, a lung nodule detection method based on deep learning is proposed for thoracic MR images. With parameter optimizing, spatial three-channel input construction, and transfer learning, a faster R-convolution neural network (CNN) is designed to locate the lung nodule region. Then, a false positive (FP) reduction scheme based on anatomical characteristics is designed to reduce FPs and preserve the true nodule. The proposed method is tested on 142 T2-weighted MR scans from the First Affiliated Hospital of Guangzhou Medical University. The sensitivity of the proposed method is 85.2% with 3.47 FPs per scan. The experimental results demonstrate that the designed faster R-CNN network and the FP reduction scheme are effective in the lung nodule detection and the FP reduction for MR images.

We are developing a computer-aided detection system to assist radiologists in the detection of lung nodules on thoracic computed tomography (CT) images. The purpose of this study was to improve the false-positive (FP) reduction stage of our algorithm by developing features that extract three-dimensional (3D) shape information from volumes of interest identified in the prescreening stage. We formulated 3D gradient field descriptors, and derived 19 gradient field features from their statistics. Six ellipsoid features were obtained by computing the lengths and the length ratios of the principal axes of an ellipsoid fitted to a segmented object. Both the gradient field features and the ellipsoid features were designed to distinguish spherical objects such as lung nodules from elongated objects such as vessels. The FP reduction performance in this new 25-dimensional feature space was compared to the performance in a 19-dimensional space that consisted of features extracted using previously developed methods. The performance in the 44-dimensional combined feature space was also evaluated. Linear discriminant analysis with stepwise feature selection was used for classification. The parameters used for feature selection were optimized using the simplex algorithm. Training and testing were performed using a leave-one-patient-out scheme. The FP reduction performances in different feature spaces were evaluated by using the area Az under the receiver operating characteristic curve and the number of FPs per CT section at a given sensitivity as accuracy measures. Our data set consisted of 82 CT scans (3551 axial sections) from 56 patients with section thickness ranging from 1.0 to 2.5 mm. Our prescreening algorithm detected 111 of the 116 solid nodules (nodule size: 3.0-30.6 mm) marked by experienced thoracic radiologists. The test Az values were 0.95 +/- 0.01, 0.88 +/- 0.02, and 0.94 +/- 0.01 in the new, previous, and combined feature spaces, respectively. The number of FPs per section at 80% sensitivity in these three feature spaces were 0.37, 1.61, and 0.34, respectively. The improvement in the test Az with the 25 new features was statistically significant (p<0.0001) compared to that with the previous 19 features alone.

Lung cancer is the leading cause of cancer related deaths in the world. The survival rate can be improved if the presence of lung nodules are detected early. This has also led to more focus being given to computer aided detection (CAD) and diagnosis of lung nodules. The arbitrariness of shape, size and texture of lung nodules is a challenge to be faced when developing these detection systems. In the proposed work we use convolutional neural networks to learn the features for nodule detection, replacing the traditional method of handcrafting features like geometric shape or texture. Our network uses the DetectNet architecture based on YOLO (You Only Look Once) to detect the nodules in CT scans of lung. In this architecture, object detection is treated as a regression problem with a single convolutional network simultaneously predicting multiple bounding boxes and class probabilities for those boxes. By performing training using chest CT scans from Lung Image Database Consortium (LIDC), NVIDIA DIGITS and Caffe deep learning framework, we show that nodule detection using this single neural network can result in reasonably low false positive rates with high sensitivity and precision.

ObjectiveTo evaluate the effect of computer-aided detection (CAD) system on observer performance in the detection of malignant lung nodules on chest radiograph.Materials and MethodsTwo hundred chest radiographs (100 normal and 100 abnormal with malignant solitary lung nodules) were evaluated. With CT and histological confirmation serving as a reference, the mean nodule size was 15.4 mm (range, 7-20 mm). Five chest radiologists and five radiology residents independently interpreted both the original radiographs and CAD output images using the sequential testing method. The performances of the observers for the detection of malignant nodules with and without CAD were compared using the jackknife free-response receiver operating characteristic analysis.ResultsFifty-nine nodules were detected by the CAD system with a false positive rate of 1.9 nodules per case. The detection of malignant lung nodules significantly increased from 0.90 to 0.92 for a group of observers, excluding one first-year resident (p = 0.04). When lowering the confidence score was not allowed, the average figure of merit also increased from 0.90 to 0.91 (p = 0.04) for all observers after a CAD review. On average, the sensitivities with and without CAD were 87% and 84%, respectively; the false positive rates per case with and without CAD were 0.19 and 0.17, respectively. The number of additional malignancies detected following true positive CAD marks ranged from zero to seven for the various observers.ConclusionThe CAD system may help improve observer performance in detecting malignant lung nodules on chest radiographs and contribute to a decrease in missed lung cancer.

Damage in structures can negatively affect their functionality and safety and leads to failure. Thus it is very important to monitor structures for occurrence, location and severity of damage. Structural health monitoring techniques provide information on the life expectancy of structures simultaneously detects and locates structural damage. Damage identification of structures based on vibration has always been important subjects and are being rapidly used to damage and location of structures. Artificial Neural Networks (ANNs) as a soft computing method using dynamic parameters of structures have been utilized increasingly for structural damage detection due to their excellent pattern recognition capability. Dynamic parameters of structure are easy to implement for damage identification and can directly linked to the topology of structure. This study presents the application of ANN for damage identification in steel beams using dynamic parameters. For identification of severity and location of damage, at first, five individual neural networks corresponding to mode 1 to mode 5 are considered. At the second step, a method based on neural network ensemble is proposed to combine the outcomes of the individual neural networks to a single network. Ensemble results were evaluated and discussed according to the differences between predicted output by ANN and desire data (target data) obtained from experimental modal analysis of structure.KeywordsArtificial neural networksMean square errorModal analysisDamage detectionNeural network ensemble

We examined the usefulness of an automated computerized scheme for distinction between benign and malignant solitary pulmonary nodules in chest radiographs, based on ROC analysis of radiologists’ performance. The average performance (Az = 0.817) of all radiologists with the computer output was significantly improved compared to that (Az = 0.743) without the computer output (p < 0.01). However, the performance (Az = 0.889) of the computer result was much better than these results. The effect of the computer output on radiologists’ decision-making was clearly demonstrated by the high correlation coefficient (R = 0.845) between the average change (%) in radiologists’ confidence level from without to with the computer output and the likelihood measures of malignancy (%). In other words, radiologists tended to change their confidence level to be consistent with the likelihood measure of malignancy presented. This effect tended to be greater in a group of attending radiologists (R = 0.872) compared with a group of residents (R = 0.714). These results indicated that this automated computerized scheme for distinction between benign and malignant solitary pulmonary nodules has a high potential to improve the diagnostic accuracy of radiologists’ decision-making when the radiologist uses the computer output as a “second opinion.”

Lung cancer is one of the major causes of cancer-related deaths due to its aggressive nature and delayed detections at advanced stages. Early detection of lung cancer is very important for the survival of an individual, and is a significant challenging problem. Generally, chest radiographs (X-ray) and computed tomography (CT) scans are used initially for the diagnosis of the malignant nodules; however, the possible existence of benign nodules leads to erroneous decisions. At early stages, the benign and the malignant nodules show very close resemblance to each other. In this paper, a novel deep learning-based model with multiple strategies is proposed for the precise diagnosis of the malignant nodules. Due to the recent achievements of deep convolutional neural networks (CNN) in image analysis, we have used two deep three-dimensional (3D) customized mixed link network (CMixNet) architectures for lung nodule detection and classification, respectively. Nodule detections were performed through faster R-CNN on efficiently-learned features from CMixNet and U-Net like encoder–decoder architecture. Classification of the nodules was performed through a gradient boosting machine (GBM) on the learned features from the designed 3D CMixNet structure. To reduce false positives and misdiagnosis results due to different types of errors, the final decision was performed in connection with physiological symptoms and clinical biomarkers. With the advent of the internet of things (IoT) and electro-medical technology, wireless body area networks (WBANs) provide continuous monitoring of patients, which helps in diagnosis of chronic diseases—especially metastatic cancers. The deep learning model for nodules’ detection and classification, combined with clinical factors, helps in the reduction of misdiagnosis and false positive (FP) results in early-stage lung cancer diagnosis. The proposed system was evaluated on LIDC-IDRI datasets in the form of sensitivity (94%) and specificity (91%), and better results were obatined compared to the existing methods.

The authors presented a novel system for automated nodule detection in lung CT exams. The approach is based on (1) a lung tissue segmentation preprocessing step, composed of histogram thresholding, seeded region growing, and mathematical morphology; (2) a filtering step, whose aim is the preliminary detection of candidate nodules (via 3D fast radial filtering) and estimation of their geometrical features (via scale space analysis); and (3) a false positive reduction (FPR) step, comprising a heuristic FPR, which applies thresholds based on geometrical features, and a supervised FPR, which is based on support vector machines classification, which in turn, is enhanced by a feature extraction algorithm based on maximum intensity projection processing and Zernike moments. The system was validated on 154 chest axial CT exams provided by the lung image database consortium public database. The authors obtained correct detection of 71% of nodules marked by all radiologists, with a false positive rate of 6.5 false positives per patient (FP/patient). A higher specificity of 2.5 FP/patient was reached with a sensitivity of 60%. An independent test on the ANODE09 competition database obtained an overall score of 0.310. The system shows a novel approach to the problem of lung nodule detection in CT scans: It relies on filtering techniques, image transforms, and descriptors rather than region growing and nodule segmentation, and the results are comparable to those of other recent systems in literature and show little dependency on the different types of nodules, which is a good sign of robustness.

Computer Aided Detection (CAD) systems provide a second opinion to radiologists in detecting lung cancer by providing automated analysis of the scans. The proposed CAD system consists of five processing steps: image acquisition, preprocessing, lung segmentation, nodule detection and false positive reduction. First, 400 CT scans are downloaded from the Lung Image Database Consortium (LIDC). Preprocessing is implemented using contrast stretching and enhancing. Lung segmentation and nodule detection stages are performed using a combination of region growing, thresholding and morphological operations. Each 3D structure is then subjected to tabular structure elimination to provide nodule candidates. In the false positive reduction stage, some of the basic nodule features are extracted from the training data to set thresholds for a simple rule-based classifier. The CAD achieved sensitivity of 77.77%, specificity of 69.5% and accuracy 70.53 % with an average 4.1 FPs/scan.

With the recent great advances in technology, the use of flat panel displays (FPD) is tending to increase. However, there is concern that the physical properties are not understood well enough by physicians to take advantage of the improved physical properties to assist them with diagnosis. Because displays of differing physical properties are being used in clinics, we have not seen enough research about the relationship between diagnostic ability and the physical properties of displays. The purpose of this study was to evaluate radiologists' performance in the diagnosis of nodules in chest radiographs with three kinds of flat panel displays: 3M- and 5M-pixel monochrome flat panel displays, and a 3M-pixel color flat panel display. Nine radiologists participated in this observer study. Sixty chest radiographs were used. Thirty-five images depicted malignant nodules, and 25 images were normal. We set for an 8-bit gray scale in advance. The radiologists' performance was evaluated with receiver operating characteristic (ROC) analysis by using a computer program (ROCKIT; Charles E. Metz, University of Chicago) . Observer performance for the diagnosis of nodules in chest radiographs was not significantly different using 3M- and 5M-pixel monochrome, and 3M-pixel color flat panel displays. Further investigation of frequent spatial patterns requiring higher resolutions such as the interstitial infiltrate pattern in chest images and the microcalcifications in mammographs will be needed.