The Lung Image Database Consortium (LIDC): an evaluation of radiologist variability in the identification of lung nodules on CT scans

The Challenge of Diagnosing Invasive Pulmonary Aspergillosis in Children: A Review of Existing and Emerging Tools

Invasive pulmonary aspergillosis remains a major cause of morbidity and mortality for immunocompromised children, particularly for patients with acute leukaemia and those undergoing haematopoietic stem cell transplantation. Timely diagnosis, using a combination of computed tomography (CT) imaging and microbiological testing, is key to improve prognosis, yet there are inherent challenges in this process. For CT imaging, changes in children are generally less specific than those reported in adults and recent data are limited. Respiratory sampling by either bronchoalveolar lavage or lung biopsy is recommended but is not always feasible in children, and serum biomarkers, including galactomannan, have important limitations. In this review we summarise the current paediatric data on available diagnostic tests for IPA and highlight key emerging diagnostic modalities with potential for future use.

Diagnostic performance of computed tomography features in detecting oropharyngeal squamous cell carcinoma extranodal extension

OBJECTIVES

Accurate pre-treatment imaging determination of extranodal extension (ENE) could facilitate the selection of appropriate initial therapy for HPV-positive oropharyngeal squamous cell carcinoma (HPV + OPSCC). Small studies have associated 7 CT features with ENE with varied results and agreement. This article seeks to determine the replicable diagnostic performance of these CT features for ENE.

METHODS

Five expert academic head/neck neuroradiologists from 5 institutions evaluate a single academic cancer center cohort of 75 consecutive HPV + OPSCC patients. In a web-based virtual laboratory for imaging research and education, the experts performed training on 7 published CT features associated with ENE and then independently identified the "single most (if any) suspicious" lymph node and presence/absence of each of the features. Inter-rater agreement was assessed using percentage agreement, Gwet's AC1, and Fleiss' kappa. Sensitivity, specificity, and positive and negative predictive values were calculated for each CT feature based on histologic ENE.

RESULTS

All 5 raters identified the same node in 52 cases (69%). In 15 cases (20%), at least one rater selected a node and at least one rater did not. In 8 cases (11%), all raters selected a node, but at least one rater selected a different node. Percentage agreement and Gwet's AC1 coefficients were > 0.80 for lesion identification, matted/conglomerated nodes, and central necrosis. Fleiss' kappa was always < 0.6. CT sensitivity for histologically confirmed ENE ranged 0.18-0.94, specificity 0.41-0.88, PPV 0.26-0.36, and NPV 0.78-0.96.

CONCLUSIONS

Previously described CT features appear to have poor reproducibility among expert head/neck neuroradiologists and poor predictive value for histologic ENE.

KEY POINTS

• Previously described CT imaging features appear to have poor reproducibility among expert head and neck subspecialized neuroradiologists as well as poor predictive value for histologic ENE. • Although it may still be appropriate to comment on the presence or absence of these CT features in imaging reports, the evidence indicates that caution is warranted when incorporating these features into clinical decision-making regarding the likelihood of ENE.

MS-Net: Learning to assess the malignant status of a lung nodule by a radiologist and her peers

BACKGROUND

Automatically assessing the malignant status of lung nodules based on CTscan images can help reduce the workload of radiologists while improving their diagnostic accuracy.

PURPOSE

Despite remarkable progress in the automatic diagnosis of pulmonary nodules by deep learning technologies, two significant problems remain outstanding. First, end-to-end deep learning solutions tend to neglect the empirical (semantic) features accumulated by radiologists and only rely on automatic features discovered by neural networks to provide the final diagnostic results, leading to questionable reliability, and interpretability. Second, inconsistent diagnosis between radiologists, a widely acknowledged phenomenon in clinical settings, is rarely examined and quantitatively explored by existing machine learning approaches. This paper solves these problems.

METHODS

We propose a novel deep neural network called MS-Net, which comprises two sequential modules: A feature derivation and initial diagnosis module (FDID), followed by a diagnosis refinement module (DR). Specifically, to take advantage of accumulated empirical features and discovered automatic features, the FDID model of MS-Net first derives a range of perceptible features and provides two initial diagnoses for lung nodules; then, these results are fed to the subsequent DR module to refine the diagnoses further. In addition, to fully consider the individual and panel diagnosis opinions, we propose a new loss function called collaborative loss, which can collaboratively optimize the individual and her peers' opinions to provide a more accurate diagnosis.

RESULTS

We evaluate the performance of the proposed MS-Net on the Lung Image Database Consortium image collection (LIDC-IDRI). It achieves 92.4% of accuracy, 92.9% of sensitivity, and 92.0% of specificity when panel labels are the ground truth, which is superior to other state-of-the-art diagnosis models. As a byproduct, the MS-Net can automatically derive a range of semantic features of lung nodules, increasing the interpretability of the final diagnoses.

CONCLUSIONS

The proposed MS-Net can provide an automatic and accurate diagnosis of lung nodules, meeting the need for a reliable computer-aided diagnosis system in clinical practice.

Artificial intelligence in lung cancer diagnosis and prognosis: Current application and future perspective

Lung cancer is one of the malignant tumors with the highest incidence and mortality in the world. The overall five-year survival rate of lung cancer is relatively lower than many leading cancers. Early diagnosis and prognosis of lung cancer are essential to improve the patient's survival rate. With artificial intelligence (AI) approaches widely applied in lung cancer, early diagnosis and prediction have achieved excellent performance in recent years. This review summarizes various types of AI algorithm applications in lung cancer, including natural language processing (NLP), machine learning and deep learning, and reinforcement learning. In addition, we provides evidence regarding the application of AI in lung cancer diagnostic and clinical prognosis. This review aims to elucidate the value of AI in lung cancer diagnosis and prognosis as the novel screening decision-making for the precise treatment of lung cancer patients.

Validation study of machine-learning chest radiograph software in primary and emergency medicine

AIM

To evaluate the performance of a machine learning based algorithm tool for chest radiographs (CXRs), applied to a consecutive cohort of historical clinical cases, in comparison to expert chest radiologists.

MATERIALS AND METHODS

The study comprised 1,960 consecutive CXR from primary care referrals and the emergency department (992 and 968 cases respectively), obtained in 2015 at a UK hospital. Two chest radiologists, each with >20 years of experience independently read all studies in consensus to serve as a reference standard. A chest artificial intelligence (AI) algorithm, Lunit INSIGHT CXR, was run on the CXRs, and results were correlated with those by the expert readers. The area under the receiver operating characteristic curve (AUROC) was calculated for the normal and 10 common findings: atelectasis, fibrosis, calcification, consolidation, lung nodules, cardiomegaly, mediastinal widening, pleural effusion, pneumothorax, and pneumoperitoneum.

RESULTS

The ground truth annotation identified 398 primary care and 578 emergency department datasets containing pathologies. The AI algorithm showed AUROC of 0.881-0.999 in the emergency department dataset and 0.881-0.998 in the primary care dataset. The AUROC for each of the findings between the primary care and emergency department datasets did not differ, except for pleural effusion (0.954 versus 0.988, p<0.001).

CONCLUSIONS

The AI algorithm can accurately and consistently differentiate normal from major thoracic abnormalities in both acute and non-acute settings, and can serve as a triage tool.

Recursive Training Strategy for a Deep Learning Network for Segmentation of Pathology Nuclei With Incomplete Annotation

This study developed a recursive training strategy to train a deep learning model for nuclei detection and segmentation using incomplete annotation. A dataset of 141 H&E stained breast cancer pathologic images with incomplete annotation was randomly split into training/validation set and test set of 89 and 52 images, respectively. The positive training samples were extracted at each annotated cell and augmented with affine translation. The negative training samples were selected from the non-cellular regions free of nuclei using a histogram-based semi-automatic method. A U-Net model was initially trained by minimizing a custom loss function. After the first stage of training, the trained U-Net model was applied to the images in the training set in an inference mode. The U-Net segmented objects with high quality were selected by a semi-automated method. Combining the newly selected high quality objects with the annotated nuclei and the previously generated negative samples, the U-Net model was retrained recursively until the stopping criteria were satisfied. For the 52 test images, the U-Net trained with and without using our recursive training method achieved a sensitivity of 90.3% and 85.3% for nuclei detection, respectively. For nuclei segmentation, the average Dice coefficient and average Jaccard index were 0.831±0.213 and 0.750±0.217, 0.780±0.270 and 0.697±0.264, for U-Net with and without recursive training, respectively. The improvement achieved by our proposed method was statistically significant (P < 0.05). In conclusion, our recursive training method effectively enlarged the set of annotated objects for training the deep learning model and further improved the detection and segmentation performance.

Tumor burden of lung metastases at initial staging in breast cancer patients detected by artificial intelligence as a prognostic tool for precision medicine

Background

Determination of the total number and size of all pulmonary metastases on chest CT is time-consuming and as such has been understudied as an independent metric for disease assessment. A novel artificial intelligence (AI) model may allow for automated detection, size determination, and quantification of the number of pulmonary metastases on chest CT.

Objective

To investigate the utility of a novel AI program applied to initial staging chest CT in breast cancer patients in risk assessment of mortality and survival.

Methods

Retrospective imaging data from a cohort of 226 subjects with breast cancer was assessed by the novel AI program and the results validated by blinded readers. Mean clinical follow-up was 2.5 years for outcomes including cancer-related death and development of extrapulmonary metastatic disease. AI measurements including total number of pulmonary metastases and maximum nodule size were assessed by Cox-proportional hazard modeling and adjusted survival.

Results

752 lung nodules were identified by the AI program, 689 of which were identified in 168 subjects having confirmed lung metastases (Lmet+) and 63 were identified in 58 subjects without confirmed lung metastases (Lmet-). When compared to the reader assessment, AI had a per-patient sensitivity, specificity, PPV and NPV of 0.952, 0.639, 0.878, and 0.830. Mortality in the Lmet + group was four times greater compared to the Lmet-group (p = 0.002). In a multivariate analysis, total lung nodule count by AI had a high correlation with overall mortality (OR 1.11 (range 1.07–1.15), p < 0.001) with an AUC of 0.811 (R2 = 0.226, p < 0.0001). When total lung nodule count and maximum nodule diameter were combined there was an AUC of 0.826 (R2 = 0.243, p < 0.001).

Conclusion

Automated AI-based detection of lung metastases in breast cancer patients at initial staging chest CT performed well at identifying pulmonary metastases and demonstrated strong correlation between the total number and maximum size of lung metastases with future mortality.

Clinical impact

As a component of precision medicine, AI-based measurements at the time of initial staging may improve prediction of which breast cancer patients will have negative future outcomes.

•

Automated detection software can quantify lung metastases on initial staging chest CT in breast cancer patients.

•

AI-detected lung metastases number and max diameter on CT at initial cancer staging were strong predictors of mortality.

•

AI detection and segmentation tool contributes to accurate individualized prognostication in breast cancer patients.

Radiological patterns of pulmonary fungal infection in pediatric hematology and oncology patients

Abstract Objective: To describe the radiological findings in pediatric patients with hematological or oncological diseases who also have an invasive fungal infection (IFI). Materials and Methods: This was a retrospective study of all patients with IFI admitted to a pediatric hematology and oncology hospital in Brazil between 2008 and 2014. Clinical and demographic data were collected. Chest computed tomography (CT) scans of the patients were reviewed by two independent radiologists. Results: We evaluated the chest CT scans of 40 pediatric patients diagnosed with an IFI. Twenty-seven patients (67.5%) had nodules with the halo sign, seven (17.5%) had cavities, two (5.0%) had nodules without the halo sign, and seven (17.5%) had consolidation. The patients with the halo sign and cavities were older (123 vs. 77 months of age; p = 0.03) and had less severe disease (34% vs. 73%; p = 0.04). Ten patients had a proven IFI: with Aspergillus sp. (n = 4); with Candida sp. (n = 5); or with Fusarium sp. (n = 1). Conclusion: A diagnosis of IFI should be considered in children and adolescents with risk factors and abnormal CT scans, even if the imaging findings are nonspecific.

A segmentation tool for pulmonary nodules in lung cancer screening: Testing and clinical usage

PURPOSE

With the future goal of defining a large dataset based on low-dose CT with labelled pulmonary lesions for lung cancer screening (LCS) research, the aim of this work is to propose and evaluate into a clinical context a tool for semi-automatic segmentation able to facilitate the process of labels collection from a LCS study (COSMOS, Continuous Observation of SMOking Subjects).

METHODS

Considering a preliminary set of manual annotations, a segmentation model based on a 2D-Unet was trained from scratch. Contour quality of the final 2D-Unet was assessed on an internal test set of manual annotations and on a subset of the public available LIDC dataset used as external test set. The tool for semi-automatic segmentation was then designed integrating the tested model into a Graphical User Interface. According to the opinion of two clinical users, the percentage of lesions properly contoured through the tool was quantified (Acceptance Rate, AR). The variability between segmentations derived by the two readers was estimated as mean percentage of difference (MPD) between the two sets of volumes and comparing the likelihood of malignancy derived from Volume Doubling Time (VDT).

RESULTS

Performance in test sets were found similar (DICE ~ 0.75(0.15)). Accordingly, a good mean AR (80.1%) resulted from the two readers. Variability in terms of MPD was equal to 23.6% while 2.7% was the VDTs percentage of disagreement.

CONCLUSIONS

A semi-automatic segmentation tool was developed and its applicability evaluated into a clinical context demonstrating the efficacy of the tool in facilitating the collection of labelled data.

On the performance of lung nodule detection, segmentation and classification

Computed tomography (CT) screening is an effective way for early detection of lung cancer in order to improve the survival rate of such a deadly disease. For more than two decades, image processing techniques such as nodule detection, segmentation, and classification have been extensively studied to assist physicians in identifying nodules from hundreds of CT slices to measure shapes and HU distributions of nodules automatically and to distinguish their malignancy. Thanks to new parallel computation, multi-layer convolution, nonlinear pooling operation, and the big data learning strategy, recent development of deep-learning algorithms has shown great progress in lung nodule screening and computer-assisted diagnosis (CADx) applications due to their high sensitivity and low false positive rates. This paper presents a survey of state-of-the-art deep-learning-based lung nodule screening and analysis techniques focusing on their performance and clinical applications, aiming to help better understand the current performance, the limitation, and the future trends of lung nodule analysis.

Radiomics in immuno-oncology

With the ongoing advances in imaging techniques, increasing volumes of anatomical and functional data are being generated as part of the routine clinical workflow. This surge of available imaging data coincides with increasing research in quantitative imaging, particularly in the domain of imaging features. An important and novel approach is radiomics, where high-dimensional image properties are extracted from routine medical images. The fundamental principle of radiomics is the hypothesis that biomedical images contain predictive information, not discernible to the human eye, that can be mined through quantitative image analysis. In this review, a general outline of radiomics and artificial intelligence (AI) will be provided, along with prominent use cases in immunotherapy (e.g. response and adverse event prediction) and targeted therapy (i.e. radiogenomics). While the increased use and development of radiomics and AI in immuno-oncology is highly promising, the technology is still in its early stages, and different challenges still need to be overcome. Nevertheless, novel AI algorithms are being constructed with an ever-increasing scope of applications.

Efficiency of a computer-aided diagnosis (CAD) system with deep learning in detection of pulmonary nodules on 1-mm-thick images of computed tomography

PURPOSE

To evaluate the performance of a deep learning-based computer-aided diagnosis (CAD) system at detecting pulmonary nodules on CT by comparing radiologists' readings with and without CAD.

MATERIALS AND METHODS

A total of 120 chest CT images were randomly selected from patients with suspected lung cancer. The gold standard of nodules ≥ 3 mm was established by a panel of three expert radiologists. Two less experienced radiologists read the images without and afterward with CAD system. Their reading times were recorded.

RESULTS

The radiologists' sensitivity increased from 20.9% to 38.0% with the introduction of CAD. The positive predictive value (PPV) decreased from 70.5% to 61.8%, and the F1-score increased from 32.2% to 47.0%. The sensitivity significantly increased from 13.7% to 32.4% for small nodules (3-6 mm) and from 33.3% to 47.6% for medium nodules (6-10 mm). CAD alone showed a sensitivity of 70.3%, a PPV of 57.9%, and an F1-score of 63.5%. Reading time decreased by 11.3% with the use of CAD.

CONCLUSION

CAD improved the less experienced radiologists' sensitivity in detecting pulmonary nodules of all sizes, especially including a significant improvement in the detection of clinically important-sized medium nodules (6-10 mm) as well as small nodules (3-6 mm) and reduced their reading time.

Pathologic categorization of lung nodules: Radiomic descriptors of CT attenuation distribution patterns of solid and subsolid nodules in low-dose CT

PURPOSE

Develop a quantitative image analysis method to characterize the heterogeneous patterns of nodule components for the classification of pathological categories of nodules.

MATERIALS AND METHODS

With IRB approval and permission of the National Lung Screening Trial (NLST) project, 103 subjects with low dose CT (LDCT) were used in this study. We developed a radiomic quantitative CT attenuation distribution descriptor (qADD) to characterize the heterogeneous patterns of nodule components and a hybrid model (qADD+) that combined qADD with subject demographic data and radiologist-provided nodule descriptors to differentiate aggressive tumors from indolent tumors or benign nodules with pathological categorization as reference standard. The classification performances of qADD and qADD + were evaluated and compared to the Brock and the Mayo Clinic models by analysis of the area under the receiver operating characteristic curve (AUC).

RESULTS

The radiomic features were consistently selected into qADDs to differentiate pathological invasive nodules from (1) preinvasive nodules, (2) benign nodules, and (3) the group of preinvasive and benign nodules, achieving test AUCs of 0.847 ± 0.002, 0.842 ± 0.002 and 0.810 ± 0.001, respectively. The qADD + obtained test AUCs of 0.867 ± 0.002, 0.888 ± 0.001 and 0.852 ± 0.001, respectively, which were higher than both the Brock and the Mayo Clinic models.

CONCLUSION

The pathologic invasiveness of lung tumors could be categorized according to the CT attenuation distribution patterns of the nodule components manifested on LDCT images, and the majority of invasive lung cancers could be identified at baseline LDCT scans.

Comparison of prediction models with radiological semantic features and radiomics in lung cancer diagnosis of the pulmonary nodules: a case-control study

PURPOSE

To compare the ability of radiological semantic and quantitative texture features in lung cancer diagnosis of pulmonary nodules.

MATERIALS AND METHODS

A total of N = 121 subjects with confirmed non-small-cell lung cancer were matched with 117 controls based on age and gender. Radiological semantic and quantitative texture features were extracted from CT images with or without contrast enhancement. Three different models were compared using LASSO logistic regression: "CS" using clinical and semantic variables, "T" using texture features, and "CST" using clinical, semantic, and texture variables. For each model, we performed 100 trials of fivefold cross-validation and the average receiver operating curve was accessed. The AUC of the cross-validation study (AUC_CV) was calculated together with its 95% confidence interval.

RESULTS

The AUC_CV (and 95% confidence interval) for models T, CS, and CST was 0.85 (0.71-0.96), 0.88 (0.77-0.96), and 0.88 (0.77-0.97), respectively. After separating the data into two groups with or without contrast enhancement, the AUC (without cross-validation) of the model T was 0.86 both for images with and without contrast enhancement, suggesting that contrast enhancement did not impact the utility of texture analysis.

CONCLUSIONS

The models with semantic and texture features provided cross-validated AUCs of 0.85-0.88 for classification of benign versus cancerous nodules, showing potential in aiding the management of patients.

KEY POINTS

• Pretest probability of cancer can aid and direct the physician in the diagnosis and management of pulmonary nodules in a cost-effective way. • Semantic features (qualitative features reported by radiologists to characterize lung lesions) and radiomic (e.g., texture) features can be extracted from CT images. • Input of these variables into a model can generate a pretest likelihood of cancer to aid clinical decision and management of pulmonary nodules.

End-to-end lung cancer screening with three-dimensional deep learning on low-dose chest computed tomography

With an estimated 160,000 deaths in 2018, lung cancer is the most common cause of cancer death in the United States¹. Lung cancer screening using low-dose computed tomography has been shown to reduce mortality by 20-43% and is now included in US screening guidelines^1-6. Existing challenges include inter-grader variability and high false-positive and false-negative rates^7-10. We propose a deep learning algorithm that uses a patient's current and prior computed tomography volumes to predict the risk of lung cancer. Our model achieves a state-of-the-art performance (94.4% area under the curve) on 6,716 National Lung Cancer Screening Trial cases, and performs similarly on an independent clinical validation set of 1,139 cases. We conducted two reader studies. When prior computed tomography imaging was not available, our model outperformed all six radiologists with absolute reductions of 11% in false positives and 5% in false negatives. Where prior computed tomography imaging was available, the model performance was on-par with the same radiologists. This creates an opportunity to optimize the screening process via computer assistance and automation. While the vast majority of patients remain unscreened, we show the potential for deep learning models to increase the accuracy, consistency and adoption of lung cancer screening worldwide.

The Added Value of Computer-aided Detection of Small Pulmonary Nodules and Missed Lung Cancers

Lung cancer at its earliest stage is typically manifested on computed tomography as a pulmonary nodule, which could be detected by low-dose multidetector computed tomography technology and the use of thinner collimation. Within the last 2 decades, computer-aided detection (CAD) of pulmonary nodules has been developed to meet the increasing demand for lung cancer screening computed tomography with a larger set of images per scan. This review introduced the basic techniques and then summarized the up-to-date applications of CAD systems in clinical and research programs and in the low-dose lung cancer screening trials, especially in the detection of small pulmonary nodules and missed lung cancers. Many studies have already shown that the CAD systems could increase the sensitivity and reduce the false-positive rate in the diagnosis of pulmonary nodules, especially for the small and isolated nodules. Further improvements to the current CAD schemes are needed to detect nodules accurately, particularly for subsolid nodules.

Classification of lung nodules in CT scans using three-dimensional deep convolutional neural networks with a checkpoint ensemble method

BACKGROUND

Accurately detecting and examining lung nodules early is key in diagnosing lung cancers and thus one of the best ways to prevent lung cancer deaths. Radiologists spend countless hours detecting small spherical-shaped nodules in computed tomography (CT) images. In addition, even after detecting nodule candidates, a considerable amount of effort and time is required for them to determine whether they are real nodules. The aim of this paper is to introduce a high performance nodule classification method that uses three dimensional deep convolutional neural networks (DCNNs) and an ensemble method to distinguish nodules between non-nodules.

METHODS

In this paper, we use a three dimensional deep convolutional neural network (3D DCNN) with shortcut connections and a 3D DCNN with dense connections for lung nodule classification. The shortcut connections and dense connections successfully alleviate the gradient vanishing problem by allowing the gradient to pass quickly and directly. Connections help deep structured networks to obtain general as well as distinctive features of lung nodules. Moreover, we increased the dimension of DCNNs from two to three to capture 3D features. Compared with shallow 3D CNNs used in previous studies, deep 3D CNNs more effectively capture the features of spherical-shaped nodules. In addition, we use an alternative ensemble method called the checkpoint ensemble method to boost performance.

RESULTS

The performance of our nodule classification method is compared with that of the state-of-the-art methods which were used in the LUng Nodule Analysis 2016 Challenge. Our method achieves higher competition performance metric (CPM) scores than the state-of-the-art methods using deep learning. In the experimental setup ESB-ALL, the 3D DCNN with shortcut connections and the 3D DCNN with dense connections using the checkpoint ensemble method achieved the highest CPM score of 0.910.

CONCLUSION

The result demonstrates that our method of using a 3D DCNN with shortcut connections, a 3D DCNN with dense connections, and the checkpoint ensemble method is effective for capturing 3D features of nodules and distinguishing nodules between non-nodules.

A Comparison of Lung Nodule Segmentation Algorithms: Methods and Results from a Multi-institutional Study

Tumor volume estimation, as well as accurate and reproducible borders segmentation in medical images, are important in the diagnosis, staging, and assessment of response to cancer therapy. The goal of this study was to demonstrate the feasibility of a multi-institutional effort to assess the repeatability and reproducibility of nodule borders and volume estimate bias of computerized segmentation algorithms in CT images of lung cancer, and to provide results from such a study. The dataset used for this evaluation consisted of 52 tumors in 41 CT volumes (40 patient datasets and 1 dataset containing scans of 12 phantom nodules of known volume) from five collections available in The Cancer Imaging Archive. Three academic institutions developing lung nodule segmentation algorithms submitted results for three repeat runs for each of the nodules. We compared the performance of lung nodule segmentation algorithms by assessing several measurements of spatial overlap and volume measurement. Nodule sizes varied from 29 μl to 66 ml and demonstrated a diversity of shapes. Agreement in spatial overlap of segmentations was significantly higher for multiple runs of the same algorithm than between segmentations generated by different algorithms (p < 0.05) and was significantly higher on the phantom dataset compared to the other datasets (p < 0.05). Algorithms differed significantly in the bias of the measured volumes of the phantom nodules (p < 0.05) underscoring the need for assessing performance on clinical data in addition to phantoms. Algorithms that most accurately estimated nodule volumes were not the most repeatable, emphasizing the need to evaluate both their accuracy and precision. There were considerable differences between algorithms, especially in a subset of heterogeneous nodules, underscoring the recommendation that the same software be used at all time points in longitudinal studies.

Vasculature surrounding a nodule: A novel lung cancer biomarker

PURPOSE

To investigate whether the vessels surrounding a nodule depicted on non-contrast, low-dose computed tomography (LDCT) can discriminate benign and malignant screen detected nodules.

MATERIALS AND METHODS

We collected a dataset consisting of LDCT scans acquired on 100 subjects from the Pittsburgh Lung Screening study (PLuSS). Fifty subjects were diagnosed with lung cancer and 50 subjects had suspicious nodules later proven benign. For the lung cancer cases, the location of the malignant nodule in the LDCT scans was known; while for the benign cases, the largest nodule in the LDCT scan was used in the analysis. A computer algorithm was developed to identify surrounding vessels and quantify the number and volume of vessels that were connected or near the nodule. A nonparametric receiver operating characteristic (ROC) analysis was performed based on a single nodule per subject to assess the discriminability of the surrounding vessels to provide a lung cancer diagnosis. Odds ratio (OR) were computed to determine the probability of a nodule being lung cancer based on the vessel features.

RESULTS

The areas under the ROC curves (AUCs) for vessel count and vessel volume were 0.722 (95% CI=0.616-0.811, p<0.01) and 0.676 (95% CI=0.565-0.772), respectively. The number of vessels attached to a nodule was significantly higher in the lung cancer group 9.7 (±9.6) compared to the non-lung cancer group 4.0 (±4.3) CONCLUSION: Our preliminary results showed that malignant nodules are often surrounded by more vessels compared to benign nodules, suggesting that the surrounding vessel characteristics could serve as lung cancer biomarker for indeterminate nodules detected during LDCT lung cancer screening using only the information collected during the initial visit.

Computerized detection of lung nodules through radiomics

PURPOSE

Lung cancer is a major cause of cancer deaths, and the 5-year survival rate of stage IV lung cancer patients is only 2%. However, the 5-year survival rate of stage I lung cancer patients significantly increases to 50%. As such, spiral computed tomography (CT) scans are necessary to diagnose high-risk lung cancer patients in early stages. In this study, a computer-aided detection (CAD) system with radiomics was proposed. This system could automatically detect pulmonary nodules and reduce radiologists' workloads and human errors.

METHODS

In the proposed scheme, a nodular enhancement filter was used to segment nodule candidates and extract radiomic features. A synthetic minority over-sampling technique was also applied to balance the samples, and a random forest method was utilized to distinguish between real nodules and false positive detections. The radiomics approach quantified intratumor heterogeneity and multifrequency information, which are highly correlated with lung nodules.

RESULTS

The proposed method was used to evaluate 1004 CT cases from the well-known Lung Image Database Consortium, and 88.9% sensitivity with four false positive detections per CT scan was obtained by randomly selecting 502 cases for training and 502 other cases for testing.

CONCLUSIONS

The proposed scheme yielded a high performance on the LIDC database. Therefore, the proposed scheme is possibly effective for various CT configurations used in routine diagnosis and lung cancer screening.

Automatic feature learning using multichannel ROI based on deep structured algorithms for computerized lung cancer diagnosis

This study aimed to analyze the ability of extracting automatically generated features using deep structured algorithms in lung nodule CT image diagnosis, and compare its performance with traditional computer aided diagnosis (CADx) systems using hand-crafted features. All of the 1018 cases were acquired from Lung Image Database Consortium (LIDC) public lung cancer database. The nodules were segmented according to four radiologists' markings, and 13,668 samples were generated by rotating every slice of nodule images. Three multichannel ROI based deep structured algorithms were designed and implemented in this study: convolutional neural network (CNN), deep belief network (DBN), and stacked denoising autoencoder (SDAE). For the comparison purpose, we also implemented a CADx system using hand-crafted features including density features, texture features and morphological features. The performance of every scheme was evaluated by using a 10-fold cross-validation method and an assessment index of the area under the receiver operating characteristic curve (AUC). The observed highest area under the curve (AUC) was 0.899±0.018 achieved by CNN, which was significantly higher than traditional CADx with the AUC=0.848±0.026. The results from DBN was also slightly higher than CADx, while SDAE was slightly lower. By visualizing the automatic generated features, we found some meaningful detectors like curvy stroke detectors from deep structured schemes. The study results showed the deep structured algorithms with automatically generated features can achieve desirable performance in lung nodule diagnosis. With well-tuned parameters and large enough dataset, the deep learning algorithms can have better performance than current popular CADx. We believe the deep learning algorithms with similar data preprocessing procedure can be used in other medical image analysis areas as well.

Improved image quality and diagnostic potential using ultra-high-resolution computed tomography of the lung with small scan FOV: A prospective study

The aim of this study was to assess whether CT imaging using an ultra-high-resolution CT (UHRCT) scan with a small scan field of view (FOV) provides higher image quality and helps to reduce the follow-up period compared with a conventional high-resolution CT (CHRCT) scan. We identified patients with at least one pulmonary nodule at our hospital from July 2015 to November 2015. CHRCT and UHRCT scans were conducted in all enrolled patients. Three experienced radiologists evaluated the image quality using a 5-point score and made diagnoses. The paired images were displayed side by side in a random manner and annotations of scan information were removed. The following parameters including image quality, diagnostic confidence of radiologists, follow-up recommendations and diagnostic accuracy were assessed. A total of 52 patients (62 nodules) were included in this study. UHRCT scan provides a better image quality regarding the margin of nodules and solid internal component compared to that of CHRCT (P < 0.05). Readers have higher diagnostic confidence based on the UHRCT images than of CHRCT images (P<0.05). The follow-up recommendations were significantly different between UHRCT and CHRCT images (P<0.05). Compared with the surgical pathological findings, UHRCT had a relative higher diagnostic accuracy than CHRCT (P > 0.05). These findings suggest that the UHRCT prototype scanner provides a better image quality of subsolid nodules compared to CHRCT and contributes significantly to reduce the patients' follow-up period.

Lung nodule malignancy classification using only radiologist-quantified image features as inputs to statistical learning algorithms: probing the Lung Image Database Consortium dataset with two statistical learning methods

In the assessment of nodules in CT scans of the lungs, a number of image-derived features are diagnostically relevant. Currently, many of these features are defined only qualitatively, so they are difficult to quantify from first principles. Nevertheless, these features (through their qualitative definitions and interpretations thereof) are often quantified via a variety of mathematical methods for the purpose of computer-aided diagnosis (CAD). To determine the potential usefulness of quantified diagnostic image features as inputs to a CAD system, we investigate the predictive capability of statistical learning methods for classifying nodule malignancy. We utilize the Lung Image Database Consortium dataset and only employ the radiologist-assigned diagnostic feature values for the lung nodules therein, as well as our derived estimates of the diameter and volume of the nodules from the radiologists' annotations. We calculate theoretical upper bounds on the classification accuracy that are achievable by an ideal classifier that only uses the radiologist-assigned feature values, and we obtain an accuracy of 85.74 [Formula: see text], which is, on average, 4.43% below the theoretical maximum of 90.17%. The corresponding area-under-the-curve (AUC) score is 0.932 ([Formula: see text]), which increases to 0.949 ([Formula: see text]) when diameter and volume features are included and has an accuracy of 88.08 [Formula: see text]. Our results are comparable to those in the literature that use algorithmically derived image-based features, which supports our hypothesis that lung nodules can be classified as malignant or benign using only quantified, diagnostic image features, and indicates the competitiveness of this approach. We also analyze how the classification accuracy depends on specific features and feature subsets, and we rank the features according to their predictive power, statistically demonstrating the top four to be spiculation, lobulation, subtlety, and calcification.

Lung nodule and cancer detection in computed tomography screening

Fundamental to the diagnosis of lung cancer in computed tomography (CT) scans is the detection and interpretation of lung nodules. As the capabilities of CT scanners have advanced, higher levels of spatial resolution reveal tinier lung abnormalities. Not all detected lung nodules should be reported; however, radiologists strive to detect all nodules that might have relevance to cancer diagnosis. Although medium to large lung nodules are detected consistently, interreader agreement and reader sensitivity for lung nodule detection diminish substantially as the nodule size falls below 8 to 10 mm. The difficulty in establishing an absolute reference standard presents a challenge to the reliability of studies performed to evaluate lung nodule detection. In the interest of improving detection performance, investigators are using eye tracking to analyze the effectiveness with which radiologists search CT scans relative to their ability to recognize nodules within their search path in order to determine whether strategies might exist to improve performance across readers. Beyond the viewing of transverse CT reconstructions, image processing techniques such as thin-slab maximum-intensity projections are used to substantially improve reader performance. Finally, the development of computer-aided detection has continued to evolve with the expectation that one day it will serve routinely as a tireless partner to the radiologist to enhance detection performance without significant prolongation of the interpretive process. This review provides an introduction to the current understanding of these varied issues as we enter the era of widespread lung cancer screening.

Computer-aided detection of pulmonary nodules: a comparative study using the public LIDC/IDRI database

Objectives

To benchmark the performance of state-of-the-art computer-aided detection (CAD) of pulmonary nodules using the largest publicly available annotated CT database (LIDC/IDRI), and to show that CAD finds lesions not identified by the LIDC’s four-fold double reading process.

Methods

The LIDC/IDRI database contains 888 thoracic CT scans with a section thickness of 2.5 mm or lower. We report performance of two commercial and one academic CAD system. The influence of presence of contrast, section thickness, and reconstruction kernel on CAD performance was assessed. Four radiologists independently analyzed the false positive CAD marks of the best CAD system.

Results

The updated commercial CAD system showed the best performance with a sensitivity of 82 % at an average of 3.1 false positive detections per scan. Forty-five false positive CAD marks were scored as nodules by all four radiologists in our study.

Conclusions

On the largest publicly available reference database for lung nodule detection in chest CT, the updated commercial CAD system locates the vast majority of pulmonary nodules at a low false positive rate. Potential for CAD is substantiated by the fact that it identifies pulmonary nodules that were not marked during the extensive four-fold LIDC annotation process.

Key Points

• CAD systems should be validated on public, heterogeneous databases.

• The LIDC/IDRI database is an excellent database for benchmarking nodule CAD.

• CAD can identify the majority of pulmonary nodules at a low false positive rate.

• CAD can identify nodules missed by an extensive two-stage annotation process.

Ultra-High-Resolution Computed Tomography of the Lung: Image Quality of a Prototype Scanner

Purpose

The image noise and image quality of a prototype ultra-high-resolution computed tomography (U-HRCT) scanner was evaluated and compared with those of conventional high-resolution CT (C-HRCT) scanners.

Materials and Methods

This study was approved by the institutional review board. A U-HRCT scanner prototype with 0.25 mm x 4 rows and operating at 120 mAs was used. The C-HRCT images were obtained using a 0.5 mm x 16 or 0.5 mm x 64 detector-row CT scanner operating at 150 mAs. Images from both scanners were reconstructed at 0.1-mm intervals; the slice thickness was 0.25 mm for the U-HRCT scanner and 0.5 mm for the C-HRCT scanners. For both scanners, the display field of view was 80 mm. The image noise of each scanner was evaluated using a phantom. U-HRCT and C-HRCT images of 53 images selected from 37 lung nodules were then observed and graded using a 5-point score by 10 board-certified thoracic radiologists. The images were presented to the observers randomly and in a blinded manner.

Results

The image noise for U-HRCT (100.87 ± 0.51 Hounsfield units [HU]) was greater than that for C-HRCT (40.41 ± 0.52 HU; P < .0001). The image quality of U-HRCT was graded as superior to that of C-HRCT (P < .0001) for all of the following parameters that were examined: margins of subsolid and solid nodules, edges of solid components and pulmonary vessels in subsolid nodules, air bronchograms, pleural indentations, margins of pulmonary vessels, edges of bronchi, and interlobar fissures.

Conclusion

Despite a larger image noise, the prototype U-HRCT scanner had a significantly better image quality than the C-HRCT scanners.

Development of a personalized training system using the Lung Image Database Consortium and Image Database resource Initiative Database

RATIONALE AND OBJECTIVES

The aim of this study was to develop a personalized training system using the Lung Image Database Consortium (LIDC) and Image Database resource Initiative (IDRI) Database, because collecting, annotating, and marking a large number of appropriate computed tomography (CT) scans, and providing the capability of dynamically selecting suitable training cases based on the performance levels of trainees and the characteristics of cases are critical for developing a efficient training system.

MATERIALS AND METHODS

A novel approach is proposed to develop a personalized radiology training system for the interpretation of lung nodules in CT scans using the Lung Image Database Consortium (LIDC) and Image Database Resource Initiative (IDRI) database, which provides a Content-Boosted Collaborative Filtering (CBCF) algorithm for predicting the difficulty level of each case of each trainee when selecting suitable cases to meet individual needs, and a diagnostic simulation tool to enable trainees to analyze and diagnose lung nodules with the help of an image processing tool and a nodule retrieval tool.

RESULTS

Preliminary evaluation of the system shows that developing a personalized training system for interpretation of lung nodules is needed and useful to enhance the professional skills of trainees.

CONCLUSIONS

The approach of developing personalized training systems using the LIDC/IDRL database is a feasible solution to the challenges of constructing specific training program in terms of cost and training efficiency.

Characterizing search, recognition, and decision in the detection of lung nodules on CT scans: elucidation with eye tracking

PURPOSE

To determine the effectiveness of radiologists' search, recognition, and acceptance of lung nodules on computed tomographic (CT) images by using eye tracking.

MATERIALS AND METHODS

This study was performed with a protocol approved by the institutional review board. All study subjects provided informed consent, and all private health information was protected in accordance with HIPAA. A remote eye tracker was used to record time-varying gaze paths while 13 radiologists interpreted 40 lung CT images with an average of 3.9 synthetic nodules (5-mm diameter) embedded randomly in the lung parenchyma. The radiologists' gaze volumes ( GV gaze volume s) were defined as the portion of the lung parenchyma within 50 pixels (approximately 3 cm) of all gaze points. The fraction of the total lung volume encompassed within the GV gaze volume s, the fraction of lung nodules encompassed within each GV gaze volume (search effectiveness), the fraction of lung nodules within the GV gaze volume detected by the reader (recognition-acceptance effectiveness), and overall sensitivity of lung nodule detection were measured.

RESULTS

Detected nodules were within 50 pixels of the nearest gaze point for 990 of 992 correct detections. On average, radiologists searched 26.7% of the lung parenchyma in 3 minutes and 16 seconds and encompassed between 86 and 143 of 157 nodules within their GV gaze volume s. Once encompassed within their GV gaze volume , the average sensitivity of nodule recognition and acceptance ranged from 47 of 100 nodules to 103 of 124 nodules (sensitivity, 0.47-0.82). Overall sensitivity ranged from 47 to 114 of 157 nodules (sensitivity, 0.30-0.73) and showed moderate correlation (r = 0.62, P = .02) with the fraction of lung volume searched.

CONCLUSION

Relationships between reader search, recognition and acceptance, and overall lung nodule detection rate can be studied with eye tracking. Radiologists appear to actively search less than half of the lung parenchyma, with substantial interreader variation in volume searched, fraction of nodules included within the search volume, sensitivity for nodules within the search volume, and overall detection rate.

Missed cancers in lung cancer screening--more than meets the eye

In lung cancer, early detection and diagnosis is of paramount importance. In 2011 the National Lung Screening Trial (NLST) demonstrated the effectiveness of computed tomography (CT) screening for lung cancer in reducing mortality, and results from other ongoing trials are expected to be published in the near future. A topic that has not been widely researched to date, however, is the cause for screening failure and missed lung cancers. In this issue of European Radiology, Scholten et al. describe a number of causes for false-negative screens. Some of the implications for CT screening and nodule management raised by this report are discussed. Key Points • Many causes exist for missed lung cancers in CT screening trials • Endobronchial structures, the hila and mediastinum are blind spots on screening CTs • The management of atypical nodular opacities on thoracic CT may be challenging.

A Performance Weighted Collaborative Filtering algorithm for personalized radiology education

Devising an accurate prediction algorithm that can predict the difficulty level of cases for individuals and then selects suitable cases for them is essential to the development of a personalized training system. In this paper, we propose a novel approach, called Performance Weighted Collaborative Filtering (PWCF), to predict the difficulty level of each case for individuals. The main idea of PWCF is to assign an optimal weight to each rating used for predicting the difficulty level of a target case for a trainee, rather than using an equal weight for all ratings as in traditional collaborative filtering methods. The assigned weight is a function of the performance level of the trainee at which the rating was made. The PWCF method and the traditional method are compared using two datasets. The experimental data are then evaluated by means of the MAE metric. Our experimental results show that PWCF outperforms the traditional methods by 8.12% and 17.05%, respectively, over the two datasets, in terms of prediction precision. This suggests that PWCF is a viable method for the development of personalized training systems in radiology education.

Enabling Ontology Based Semantic Queries in Biomedical Database Systems

There is a lack of tools to ease the integration and ontology based semantic queries in biomedical databases, which are often annotated with ontology concepts. We aim to provide a middle layer between ontology repositories and semantically annotated databases to support semantic queries directly in the databases with expressive standard database query languages. We have developed a semantic query engine that provides semantic reasoning and query processing, and translates the queries into ontology repository operations on NCBO BioPortal. Semantic operators are implemented in the database as user defined functions extended to the database engine, thus semantic queries can be directly specified in standard database query languages such as SQL and XQuery. The system provides caching management to boosts query performance. The system is highly adaptable to support different ontologies through easy customizations. We have implemented the system DBOntoLink as an open source software, which supports major ontologies hosted at BioPortal. DBOntoLink supports a set of common ontology based semantic operations and have them fully integrated with a database management system IBM DB2. The system has been deployed and evaluated with an existing biomedical database for managing and querying image annotations and markups (AIM). Our performance study demonstrates the high expressiveness of semantic queries and the high efficiency of the queries.

Area--a consistent method to evaluate pulmonary tumor size on multidetector CT imaging: an intraobserver and interobserver agreement study

OBJECTIVES

To evaluate the agreements of unidimensional, bidimensional, area and volume measurements of pulmonary tumors on multidetector computed tomography (MDCT), and to determine which method is the most reliable one.

MATERIALS AND METHODS

Thirty patients with pulmonary tumors were enrolled in this study, which referred to undergo thoracic MDCT in our hospital. Four radiologists evaluated dimensions of pulmonary tumor independently, including length, width, height, area and volume. The intraclass correlation coefficient (ICC) and concordance correlation coefficient (CCC) were both used to evaluate the variability between the repeat readings of the same scan.

RESULTS

The ICCs and CCCs of the intraobserver were both higher than interobserver's (ICC intra vs. inter: 0.984 vs. 0.947 and CCC intra vs. inter: 0.993 vs. 0.943). Area of intraobserver ICC (ICC=0.992, P<.001) and CCC (CCC=0.997, P<.001) both had the best agreements of the six methods. Among the interobserver ICCs and CCCs, area (ICC=0.981, P<.001 and CCC=0.982, P<.001) was also the best of the six methods.

CONCLUSIONS

Area measurement on MDCT is the most reproducible method that measures tumor dimension accurately.

Scanners and drillers: characterizing expert visual search through volumetric images

Modern imaging methods like computed tomography (CT) generate 3-D volumes of image data. How do radiologists search through such images? Are certain strategies more efficient? Although there is a large literature devoted to understanding search in 2-D, relatively little is known about search in volumetric space. In recent years, with the ever-increasing popularity of volumetric medical imaging, this question has taken on increased importance as we try to understand, and ultimately reduce, errors in diagnostic radiology. In the current study, we asked 24 radiologists to search chest CTs for lung nodules that could indicate lung cancer. To search, radiologists scrolled up and down through a "stack" of 2-D chest CT "slices." At each moment, we tracked eye movements in the 2-D image plane and coregistered eye position with the current slice. We used these data to create a 3-D representation of the eye movements through the image volume. Radiologists tended to follow one of two dominant search strategies: "drilling" and "scanning." Drillers restrict eye movements to a small region of the lung while quickly scrolling through depth. Scanners move more slowly through depth and search an entire level of the lung before moving on to the next level in depth. Driller performance was superior to the scanners on a variety of metrics, including lung nodule detection rate, percentage of the lung covered, and the percentage of search errors where a nodule was never fixated.

Identifying resistance mechanisms against five tyrosine kinase inhibitors targeting the ERBB/RAS pathway in 45 cancer cell lines

Because of the low overall response rates of 10–47% to targeted cancer therapeutics, there is an increasing need for predictive biomarkers. We aimed to identify genes predicting response to five already approved tyrosine kinase inhibitors. We tested 45 cancer cell lines for sensitivity to sunitinib, erlotinib, lapatinib, sorafenib and gefitinib at the clinically administered doses. A resistance matrix was determined, and gene expression profiles of the subsets of resistant vs. sensitive cell lines were compared. Triplicate gene expression signatures were obtained from the caArray project. Significance analysis of microarrays and rank products were applied for feature selection. Ninety-five genes were also measured by RT-PCR. In case of four sunitinib resistance associated genes, the results were validated in clinical samples by immunohistochemistry. A list of 63 top genes associated with resistance against the five tyrosine kinase inhibitors was identified. Quantitative RT-PCR analysis confirmed 45 of 63 genes identified by microarray analysis. Only two genes (ANXA3 and RAB25) were related to sensitivity against more than three inhibitors. The immunohistochemical analysis of sunitinib-treated metastatic renal cell carcinomas confirmed the correlation between RAB17, LGALS8, and EPCAM and overall survival. In summary, we determined predictive biomarkers for five tyrosine kinase inhibitors, and validated sunitinib resistance biomarkers by immunohistochemistry in an independent patient cohort.

Lung cancer with scattered consolidation: detection of new independent radiological category of peripheral lung cancer on thin-section computed tomography

OBJECTIVES

Ground glass opacity (GGO) on thin-section computed tomography (CT) has been reported to be a favourable prognostic marker in lung cancer, and the size or area of GGO is commonly used for preoperative evaluation. However, it can sometimes be difficult to evaluate the status of GGO.

METHODS

A retrospective study was conducted on 572 consecutive patients with resected lung cancer of clinical stage IA between 2004 and 2011. All patients underwent preoperative CT and their radiological findings were reviewed. The areas of consolidation and GGO were evaluated for all lung cancers. Lung cancers were divided into three categories on the basis of the status of GGO: GGO, part solid and pure solid. Lung cancers in which it was difficult to measure GGO were selected and their clinicopathological features were investigated.

RESULTS

Seventy-one (12.4%) patients had lung cancer in whom it was difficult to measure GGO. In all these cases, consolidation and GGO were not easily measured because of their scattered distribution. In this cohort, nodal metastases were not observed at all. The frequency of other pathological factors, such as lymphatic and/or vascular invasion, was significantly lower (P < 0.0001).

CONCLUSIONS

This new category of lung cancer with scattered consolidation on thin-section CT scan tended to be pathologically less invasive. When lung cancer has GGO and is difficult to measure because of a scattered distribution, its prognosis could be favourable regardless of the area of GGO. This new category could be useful for the preoperative evaluation of lung cancer.

Quantifying the margin sharpness of lesions on radiological images for content-based image retrieval

PURPOSE

To develop a method to quantify the margin sharpness of lesions on CT and to evaluate it in simulations and CT scans of liver and lung lesions.

METHODS

The authors computed two attributes of margin sharpness: the intensity difference between a lesion and its surroundings, and the sharpness of the intensity transition across the lesion boundary. These two attributes were extracted from sigmoid curves fitted along lines automatically drawn orthogonal to the lesion margin. The authors then represented the margin characteristics for each lesion by a feature vector containing histograms of these parameters. The authors created 100 simulated CT scans of lesions over a range of intensity difference and margin sharpness, and used the concordance correlation between the known parameter and the corresponding computed feature as a measure of performance. The authors also evaluated their method in 79 liver lesions (44 patients: 23 M, 21 F, mean age 61) and 58 lung nodules (57 patients: 24 M, 33 F, mean age 66). The methodology presented takes into consideration the boundary of the liver and lung during feature extraction in clinical images to ensure that the margin feature do not get contaminated by anatomy other than the normal organ surrounding the lesions. For evaluation in these clinical images, the authors created subjective independent reference standards for pairwise margin sharpness similarity in the liver and lung cohorts, and compared rank orderings of similarity used using our sharpness feature to that expected from the reference standards using mean normalized discounted cumulative gain (NDCG) over all query images. In addition, the authors compared their proposed feature with two existing techniques for lesion margin characterization using the simulated and clinical datasets. The authors also evaluated the robustness of their features against variations in delineation of the lesion margin by simulating five types of deformations of the lesion margin. Equivalence across deformations was assessed using Schuirmann's paired two one-sided tests.

RESULTS

In simulated images, the concordance correlation between measured gradient and actual gradient was 0.994. The mean (s.d.) and standard deviation NDCG score for the retrieval of K images, K = 5, 10, and 15, were 84% (8%), 85% (7%), and 85% (7%) for CT images containing liver lesions, and 82% (7%), 84% (6%), and 85% (4%) for CT images containing lung nodules, respectively. The authors' proposed method outperformed the two existing margin characterization methods in average NDCG scores over all K, by 1.5% and 3% in datasets containing liver lesion, and 4.5% and 5% in datasets containing lung nodules. Equivalence testing showed that the authors' feature is more robust across all margin deformations (p < 0.05) than the two existing methods for margin sharpness characterization in both simulated and clinical datasets.

CONCLUSIONS

The authors have described a new image feature to quantify the margin sharpness of lesions. It has strong correlation with known margin sharpness in simulated images and in clinical CT images containing liver lesions and lung nodules. This image feature has excellent performance for retrieving images with similar margin characteristics, suggesting potential utility, in conjunction with other lesion features, for content-based image retrieval applications.

Automatic annotation of radiological observations in liver CT images

We aim to predict radiological observations using computationally-derived imaging features extracted from computed tomography (CT) images. We created a dataset of 79 CT images containing liver lesions identified and annotated by a radiologist using a controlled vocabulary of 76 semantic terms. Computationally-derived features were extracted describing intensity, texture, shape, and edge sharpness. Traditional logistic regression was compared to L(1)-regularized logistic regression (LASSO) in order to predict the radiological observations using computational features. The approach was evaluated by leave one out cross-validation. Informative radiological observations such as lesion enhancement, hypervascular attenuation, and homogeneous retention were predicted well by computational features. By exploiting relationships between computational and semantic features, this approach could lead to more accurate and efficient radiology reporting.

Illustration of the obstacles in computerized lung segmentation using examples

PURPOSE

Automated lung volume segmentation is often a preprocessing step in quantitative lung computed tomography (CT) image analysis. The objective of this study is to identify the obstacles in computerized lung volume segmentation and illustrate those explicitly using real examples. Awareness of these "difficult" cases may be helpful for the development of a robust and consistent lung segmentation algorithm.

METHODS

We collected a large diverse dataset consisting of 2768 chest CT examinations acquired on 2292 subjects from various sources. These examinations cover a wide range of diseases, including lung cancer, chronic obstructive pulmonary disease, human immunodeficiency virus, pulmonary embolism, pneumonia, asthma, and interstitial lung disease (ILD). The CT acquisition protocols, including dose, scanners, and reconstruction kernels, vary significantly. After the application of a "neutral" thresholding-based approach to the collected CT examinations in a batch manner, the failed cases were subjectively identified and classified into different subgroups.

RESULTS

Totally, 121 failed examinations are identified, corresponding to a failure ratio of 4.4%. These failed cases are summarized as 11 different subgroups, which is further classified into 3 broad categories: (1) failure caused by diseases, (2) failure caused by anatomy variability, and (3) failure caused by external factors. The failure percentages in these categories are 62.0%, 32.2%, and 5.8%, respectively.

CONCLUSIONS

The presence of specific lung diseases (e.g., pulmonary nodules, ILD, and pneumonia) is the primary issue in computerized lung segmentation. The segmentation failures caused by external factors and anatomy variety are relatively low but unavoidable in practice. It is desirable to develop robust schemes to handle these issues in a single pass when a large number of CT examinations need to be analyzed.

Extraction of lesion-partitioned features and retrieval of contrast-enhanced liver images

The most critical step in grayscale medical image retrieval systems is feature extraction. Understanding the interrelatedness between the characteristics of lesion images and corresponding imaging features is crucial for image training, as well as for features extraction. A feature-extraction algorithm is developed based on different imaging properties of lesions and on the discrepancy in density between the lesions and their surrounding normal liver tissues in triple-phase contrast-enhanced computed tomographic (CT) scans. The algorithm includes mainly two processes: (1) distance transformation, which is used to divide the lesion into distinct regions and represents the spatial structure distribution and (2) representation using bag of visual words (BoW) based on regions. The evaluation of this system based on the proposed feature extraction algorithm shows excellent retrieval results for three types of liver lesions visible on triple-phase scans CT images. The results of the proposed feature extraction algorithm show that although single-phase scans achieve the average precision of 81.9%, 80.8%, and 70.2%, dual- and triple-phase scans achieve 86.3% and 88.0%.

Informatics in radiology: Measuring and improving quality in radiology: meeting the challenge with informatics

Quality is becoming a critical issue for radiology. Measuring and improving quality is essential not only to ensure optimum effectiveness of care and comply with increasing regulatory requirements, but also to combat current trends leading to commoditization of radiology services. A key challenge to implementing quality improvement programs is to develop methods to collect knowledge related to quality care and to deliver that knowledge to practitioners at the point of care. There are many dimensions to quality in radiology that need to be measured, monitored, and improved, including examination appropriateness, procedure protocol, accuracy of interpretation, communication of imaging results, and measuring and monitoring performance improvement in quality, safety, and efficiency. Informatics provides the key technologies that can enable radiologists to measure and improve quality. However, few institutions recognize the opportunities that informatics methods provide to improve safety and quality. The information technology infrastructure in most hospitals is limited, and they have suboptimal adoption of informatics techniques. Institutions can tackle the challenges of assessing and improving quality in radiology by means of informatics.

Enabling Ontology Based Semantic Queries in Biomedical Database Systems

While current biomedical ontology repositories offer primitive query capabilities, it is difficult or cumbersome to support ontology based semantic queries directly in semantically annotated biomedical databases. The problem may be largely attributed to the mismatch between the models of the ontologies and the databases, and the mismatch between the query interfaces of the two systems. To fully realize semantic query capabilities based on ontologies, we develop a system DBOntoLink to provide unified semantic query interfaces by extending database query languages. With DBOntoLink, semantic queries can be directly and naturally specified as extended functions of the database query languages without any programming needed. DBOntoLink is adaptable to different ontologies through customizations and supports major biomedical ontologies hosted at the NCBO BioPortal. We demonstrate the use of DBOntoLink in a real world biomedical database with semantically annotated medical image annotations.

A novel computer-aided lung nodule detection system for CT images

PURPOSE

The paper presents a complete computer-aided detection (CAD) system for the detection of lung nodules in computed tomography images. A new mixed feature selection and classification methodology is applied for the first time on a difficult medical image analysis problem.

METHODS

The CAD system was trained and tested on images from the publicly available Lung Image Database Consortium (LIDC) on the National Cancer Institute website. The detection stage of the system consists of a nodule segmentation method based on nodule and vessel enhancement filters and a computed divergence feature to locate the centers of the nodule clusters. In the subsequent classification stage, invariant features, defined on a gauge coordinates system, are used to differentiate between real nodules and some forms of blood vessels that are easily generating false positive detections. The performance of the novel feature-selective classifier based on genetic algorithms and artificial neural networks (ANNs) is compared with that of two other established classifiers, namely, support vector machines (SVMs) and fixed-topology neural networks. A set of 235 randomly selected cases from the LIDC database was used to train the CAD system. The system has been tested on 125 independent cases from the LIDC database.

RESULTS

The overall performance of the fixed-topology ANN classifier slightly exceeds that of the other classifiers, provided the number of internal ANN nodes is chosen well. Making educated guesses about the number of internal ANN nodes is not needed in the new feature-selective classifier, and therefore this classifier remains interesting due to its flexibility and adaptability to the complexity of the classification problem to be solved. Our fixed-topology ANN classifier with 11 hidden nodes reaches a detection sensitivity of 87.5% with an average of four false positives per scan, for nodules with diameter greater than or equal to 3 mm. Analysis of the false positive items reveals that a considerable proportion (18%) of them are smaller nodules, less than 3 mm in diameter.

CONCLUSIONS

A complete CAD system incorporating novel features is presented, and its performance with three separate classifiers is compared and analyzed. The overall performance of our CAD system equipped with any of the three classifiers is well with respect to other methods described in literature.

VERITAS: COMBINING EXPERT OPINIONS WITHOUT LABELED DATA

We consider a variation of the problem of combining expert opinions for the situation in which there is no ground truth to use for training. Even though we do not have labeled data, the goal of this work is quite different from an unsupervised learning problem in which the goal is to cluster the data. Our work is motivated by the application of segmenting a lung nodule in a computed tomography (CT) scan of the human chest. The lack of a gold standard of truth is a critical problem in medical imaging. A variety of experts, both human and computer algorithms, are available that can mark which voxels are part of a nodule. The question is, how to combine these expert opinions to estimate the unknown ground truth. We present the Veritas algorithm that predicts the underlying label using the knowledge in the expert opinions even without the benefit of any labeled data for training. We evaluate Veritas using artificial data and real CT images to which synthetic nodules have been added, providing a known ground truth.

CT patterns of fungal pulmonary infections of the lung: comparison of standard-dose and simulated low-dose CT

PURPOSE

To assess the effect of radiation dose reduction on the appearance and visual quantification of specific CT patterns of fungal infection in immuno-compromised patients.

MATERIALS AND METHODS

Raw data of thoracic CT scans (64 × 0.75 mm, 120 kVp, 300 reference mAs) from 41 consecutive patients with clinical suspicion of pulmonary fungal infection were collected. In 32 patients fungal infection could be proven (median age of 55.5 years, range 35-83). A total of 267 cuboids showing CT patterns of fungal infection and 27 cubes having no disease were reconstructed at the original and 6 simulated tube currents of 100, 40, 30, 20, 10, and 5 reference mAs. Eight specific fungal CT patterns were analyzed by three radiologists: 76 ground glass opacities, 42 ground glass nodules, 51 mixed, part solid, part ground glass nodules, 36 solid nodules, 5 lobulated nodules, 6 spiculated nodules, 14 cavitary nodules, and 37 foci of air-space disease. The standard of reference was a consensus subjective interpretation by experts whom were not readers in the study.

RESULTS

The mean sensitivity and standard deviation for detecting pathological cuboids/disease using standard dose CT was 0.91 ± 0.07. Decreasing dose did not affect sensitivity significantly until the lowest dose level of 5 mAs (0.87 ± 0.10, p=0.012). Nodular pattern discrimination was impaired below the dose level of 30 reference mAs: specificity for fungal 'mixed nodules' decreased significantly at 20, 10 and 5 reference mAs (p<0.05). At lower dose levels, classification drifted from 'solid' to 'mixed nodule', although no lesion was missed.

CONCLUSION

Our simulation data suggest that tube current levels can be reduced from 300 to 30 reference mAs without impairing the diagnostic information of specific CT patterns of pulmonary fungal infections.