Pulmonary Nodule Detection with Low-Dose CT of the Lung: Agreement Among Radiologists

Application of magnetic resonance imaging radiomics in endometrial cancer: a systematic review and meta-analysis

PURPOSE

We aimed to systematically assess the methodological quality and clinical potential application of published magnetic resonance imaging (MRI)-based radiomics studies about endometrial cancer (EC).

METHODS

Studies of EC radiomics analyses published between 1 January 2000 and 19 March 2023 were extracted, and their methodological quality was evaluated using the radiomics quality score (RQS) and Quality Assessment of Diagnostic Accuracy Studies 2 (QUADAS-2). Pairwise correlation analyses and separate meta-analyses of studies exploring differential diagnoses and risk prediction were also performed.

RESULTS

Forty-five studies involving 3 aims were included. The mean RQS was 13.77 (range: 9-22.5); publication bias was observed in the areas of 'index test' and 'flow and timing'. A high RQS was significantly associated with therapy selection-aimed studies, low QUADAS-2 risk, recent publication year, and high-performance metrics. Raw data from 6 differential diagnosis and 34 risk prediction models were subjected to meta-analysis, revealing diagnostic odds ratios of 23.81 (95% confidence interval [CI] 8.48-66.83) and 18.23 (95% CI 13.68-24.29), respectively.

CONCLUSION

The methodological quality of radiomics studies involving patients with EC is unsatisfactory. However, MRI-based radiomics analyses showed promising utility in terms of differential diagnosis and risk prediction.

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.

Prospective Study of Spatial Distribution of Missed Lung Nodules by Readers in CT Lung Screening Using Computer-assisted Detection

PURPOSE

To evaluate the spatial patterns of missed lung nodules in a real-life routine screening environment.

MATERIALS AND METHODS

In a screening institute, 4,822 consecutive adults underwent chest CT, and each image set was independently interpreted by two radiologists in three steps: (1) independently interpreted without computer-assisted detection (CAD) software, (2) independently referred to the CAD results, (3) determined by the consensus of the two radiologists. The locations of nodules and the detection performance data were semi-automatically collected using a CAD server integrated into the reporting system. Fisher's exact test was employed for evaluating findings in different lung divisions. Probability maps were drawn to illustrate the spatial distribution of radiologists' missed nodules.

RESULTS

Radiologists significantly tended to miss lung nodules in the bilateral hilar divisions (p < 0.01). Some radiologists had their own spatial pattern of missed lung nodules.

CONCLUSION

Radiologists tend to miss lung nodules present in the hilar regions significantly more often than in the rest of the lung.

The radiologist's role in lung cancer screening

Lung cancer is still the deadliest cancer in men and women worldwide. This high mortality is related to diagnosis in advanced stages, when curative treatment is no longer an option. Large randomized controlled trials have shown that lung cancer screening (LCS) with low-dose computed tomography (CT) can detect lung cancers at earlier stages and reduce lung cancer-specific mortality. The recent publication of the significant reduction of cancer-related mortality by 26% in the Dutch-Belgian NELSON LCS trial has increased the likelihood that implementation of LCS in Europe will move forward. Radiologists are important stakeholders in numerous aspects of the LCS pathway. Their role goes beyond nodule detection and nodule management. Being part of a multidisciplinary team, radiologists are key players in numerous aspects of implementation of a high quality LCS program. In this non-systematic review we discuss the multifaceted role of radiologists in LCS.

Interobserver variability in Lung CT Screening Reporting and Data System categorisation in subsolid nodule-enriched lung cancer screening CTs

OBJECTIVES

To assess interobserver agreement in Lung CT Screening Reporting and Data System (Lung-RADS) categorisation in subsolid nodule-enriched low-dose screening CTs.

METHODS

A retrospective review of low-dose screening CT reports from 2013 to 2017 using keyword searches for subsolid nodules identified 54 baseline CT scans. With an additional 108 negative screening CT scans, a total of 162 CT scans were categorised according to the Lung-RADS by two fellowship-trained thoracic radiologists in consensus. We randomly selected 20, 20, 10, and 10 scans from categories 1/2, 3, 4A, and 4B CT scans, respectively, to ensure balanced category representation. Five radiologists classified the 60 CT scans into Lung-RADS categories. The frequencies of concordance and minor and major discordance were calculated, with major discordance defined as at least 6 months of management discrepancy. We used Cohen's κ statistics to analyse reader agreement.

RESULTS

An average of 60.3% (181 of 300) of all cases and 45.0% (90 of 200) of positive screens were correctly categorised. The minor and major discordance rates were 12.3% and 27.3% overall and 18.5% and 36.5% in positive screens, respectively. The concordance rate was significantly higher among experienced thoracic radiologists. Overall, the interobserver agreement was moderate (mean κ, 0.45; 95% confidence interval: 0.40-0.51). The proportion of part-solid risk-dominant nodules was significantly higher in cases with low rates of accurate categorisation.

CONCLUSION

This retrospective study observed variable accuracy and moderate interobserver agreement in radiologist categorisation of subsolid nodules in screening CTs. This inconsistency may affect management recommendations for lung cancer screening.

KEY POINTS

• Diagnostic performance for Lung-RADS categorisation is variable among radiologists with fair to moderate interobserver agreement in subsolid nodule-enriched CT scans. • Experienced thoracic radiologists showed more accurate and consistent Lung-RADS categorisation than radiology residents. • The relative abundance of part-solid nodules was a potential factor related to increased disagreement in Lung-RADS categorisation.

Dosimetry and Comparison between Different CT Protocols (Low Dose, Ultralow Dose, and Conventional CT) for Lung Nodules' Detection in a Phantom

Background

The effects of dose reduction in lung nodule detection need better understanding.

Purpose

To compare the detection rate of simulated lung nodules in a chest phantom using different computed tomography protocols, low dose (LD), ultralow dose (ULD), and conventional (CCT), and to quantify their respective amount of radiation.

Materials and Methods

A chest phantom containing 93 simulated lung nodules was scanned using five different protocols: ULD (80 kVp/30 mA), LD A (120 kVp/20 mA), LD B (100 kVp/30 mA), LD C (120 kVp/30 mA), and CCT (120 kVp/automatic mA). Four chest radiologists analyzed a selected image from each protocol and registered in diagrams the nodules they detected. Kruskal-Wallis and McNemar's tests were performed to determine the difference in nodule detection. Equivalent doses were estimated by placing thermoluminescent dosimeters on the surface and inside the phantom.

Results

There was no significant difference in lung nodules' detection when comparing ULD and LD protocols (p=0.208 to p=1.000), but there was a significant difference when comparing each one of those against CCT (p < 0.001). The detection rate of nodules with CT attenuation values lower than -600 HU was also different when comparing all protocols against CCT (p < 0.001 to p=0.007). There was at least moderate agreement between observers in all protocols (κ-value >0.41). Equivalent dose values ranged from 0.5 to 9 mSv.

Conclusion

There is no significant difference in simulated lung nodules' detection when comparing ULD and LD protocols, but both differ from CCT, especially when considering lower-attenuating nodules.

Diagnostic accuracy of low-dose and ultra-low-dose CT in detection of chest pathology: a systematic review

PURPOSE

Studies have evaluated imaging modalities with a lower radiation dose than standard-dose CT (SD-CT) for chest examination. This systematic review aimed to summarize evidence on diagnostic accuracy of these modalities - low-dose and ultra-low-dose CT (LD- and ULD-CT) - for chest pathology.

METHOD

Ovid-MEDLINE, Ovid-EMBASE and the Cochrane Library were systematically searched April 29th-30th, 2019 and screened by two reviewers. Studies on diagnostic accuracy were included if they defined their index tests as 'LD-CT', 'Reduced-dose CT' or 'ULD-CT' and had SD-CT as reference standard. Risk of bias was evaluated on study level using the Quality Assessment of Diagnostic Accuracy Studies-2. A narrative synthesis was conducted to compare the diagnostic accuracy measurements.

RESULTS

Of the 4257 studies identified, 18 were eligible for inclusion. SD-CT (3.17 ± 1.47 mSv) was used as reference standard in all studies to evaluate diagnostic accuracy of LD- (1.22 ± 0.34 mSv) and ULD-CT (0.22 ± 0.05 mSv), respectively. LD-CT had high sensitivities for detection of bronchiectasis (82-96%), honeycomb (75-100%), and varying sensitivities for nodules (63-99%) and ground glass opacities (GGO) (77-91%). ULD-CT had high sensitivities for GGO (93-100%), pneumothorax (100%), consolidations (90-100%), and varying sensitivities for nodules (60-100%) and emphysema (65-90%).

CONCLUSION

The included studies found LD-CT to have high diagnostic accuracy in detection of honeycombing and bronchiectasis and ULD-CT to have high diagnostic accuracy for pneumothorax, consolidations and GGO. Summarizing evidence on diagnostic accuracy of LD- and ULD-CT for other chest pathology was not possible due to varying outcome measures, lack of precision estimates and heterogeneous study design and methodology.

Pulmonary Textures Classification via a Multi-Scale Attention Network

Precise classification of pulmonary textures is crucial to develop a computer aided diagnosis (CAD) system of diffuse lung diseases (DLDs). Although deep learning techniques have been applied to this task, the classification performance is not satisfied for clinical requirements, since commonly-used deep networks built by stacking convolutional blocks are not able to learn discriminative feature representation to distinguish complex pulmonary textures. For addressing this problem, we design a multi-scale attention network (MSAN) architecture comprised by several stacked residual attention modules followed by a multi-scale fusion module. Our deep network can not only exploit powerful information on different scales but also automatically select optimal features for more discriminative feature representation. Besides, we develop visualization techniques to make the proposed deep model transparent for humans. The proposed method is evaluated by using a large dataset. Experimental results show that our method has achieved the average classification accuracy of 94.78% and the average f-value of 0.9475 in the classification of 7 categories of pulmonary textures. Besides, visualization results intuitively explain the working behavior of the deep network. The proposed method has achieved the state-of-the-art performance to classify pulmonary textures on high resolution CT images.

Developing global image feature analysis models to predict cancer risk and prognosis

In order to develop precision or personalized medicine, identifying new quantitative imaging markers and building machine learning models to predict cancer risk and prognosis has been attracting broad research interest recently. Most of these research approaches use the similar concepts of the conventional computer-aided detection schemes of medical images, which include steps in detecting and segmenting suspicious regions or tumors, followed by training machine learning models based on the fusion of multiple image features computed from the segmented regions or tumors. However, due to the heterogeneity and boundary fuzziness of the suspicious regions or tumors, segmenting subtle regions is often difficult and unreliable. Additionally, ignoring global and/or background parenchymal tissue characteristics may also be a limitation of the conventional approaches. In our recent studies, we investigated the feasibility of developing new computer-aided schemes implemented with the machine learning models that are trained by global image features to predict cancer risk and prognosis. We trained and tested several models using images obtained from full-field digital mammography, magnetic resonance imaging, and computed tomography of breast, lung, and ovarian cancers. Study results showed that many of these new models yielded higher performance than other approaches used in current clinical practice. Furthermore, the computed global image features also contain complementary information from the features computed from the segmented regions or tumors in predicting cancer prognosis. Therefore, the global image features can be used alone to develop new case-based prediction models or can be added to current tumor-based models to increase their discriminatory power.

Lung cancer screening: nodule identification and characterization

The accurate identification and characterization of small pulmonary nodules at low-dose CT is an essential requirement for the implementation of effective lung cancer screening. Individual reader detection performance is influenced by nodule characteristics and technical CT parameters but can be improved by training, the application of CT techniques, and by computer-aided techniques. However, the evaluation of nodule detection in lung cancer screening trials differs from the assessment of individual readers as it incorporates multiple readers, their inter-observer variability, reporting thresholds, and reflects the program accuracy in identifying lung cancer. Understanding detection and interpretation errors in screening trials aids in the implementation of lung cancer screening in clinical practice. Indeed, as CT screening moves to ever lower radiation doses, radiologists must be cognisant of new technical challenges in nodule assessment. Screen detected lung cancers demonstrate distinct morphological features from incidentally or symptomatically detected lung cancers. Hence characterization of screen detected nodules requires an awareness of emerging concepts in early lung cancer appearances and their impact on radiological assessment and malignancy prediction models. Ultimately many nodules remain indeterminate, but further imaging evaluation can be appropriate with judicious utilization of contrast enhanced CT or MRI techniques or functional evaluation by PET-CT.

Automatic detection of pulmonary nodules in CT images by incorporating 3D tensor filtering with local image feature analysis

Computer-aided detection (CAD) technology has been developed and demonstrated its potential to assist radiologists in detecting pulmonary nodules especially at an early stage. In this paper, we present a novel scheme for automatic detection of pulmonary nodules in CT images based on a 3D tensor filtering algorithm and local image feature analysis. We first apply a series of preprocessing steps to segment the lung volume and generate the isotropic volumetric CT data. Next, a unique 3D tensor filtering approach and local image feature analysis are used to detect nodule candidates. A 3D level set segmentation method is used to correct and refine the boundaries of nodule candidates subsequently. Then, we extract the features of the detected candidates and select the optimal features by using a CFS (Correlation Feature Selection) subset evaluator attribute selection method. Finally, a random forest classifier is trained to classify the detected candidates. The performance of this CAD scheme is validated using two datasets namely, the LUNA16 (Lung Nodule Analysis 2016) database and the ANODE09 (Automatic Nodule Detection 2009) database. By applying a 10-fold cross-validation method, the CAD scheme yielded a sensitivity of 79.3% at an average of 4 false positive detections per scan (FP/Scan) for the former dataset, and a sensitivity of 84.62% and 2.8 FP/Scan for the latter dataset, respectively. Our detection results show that the use of 3D tensor filtering algorithm combined with local image feature analysis constitutes an effective approach to detect pulmonary nodules.

Application of the 3D slicer chest imaging platform segmentation algorithm for large lung nodule delineation

PURPOSE

Accurate segmentation of lung nodules is crucial in the development of imaging biomarkers for predicting malignancy of the nodules. Manual segmentation is time consuming and affected by inter-observer variability. We evaluated the robustness and accuracy of a publically available semiautomatic segmentation algorithm that is implemented in the 3D Slicer Chest Imaging Platform (CIP) and compared it with the performance of manual segmentation.

METHODS

CT images of 354 manually segmented nodules were downloaded from the LIDC database. Four radiologists performed the manual segmentation and assessed various nodule characteristics. The semiautomatic CIP segmentation was initialized using the centroid of the manual segmentations, thereby generating four contours for each nodule. The robustness of both segmentation methods was assessed using the region of uncertainty (δ) and Dice similarity index (DSI). The robustness of the segmentation methods was compared using the Wilcoxon-signed rank test (pWilcoxon<0.05). The Dice similarity index (DSIAgree) between the manual and CIP segmentations was computed to estimate the accuracy of the semiautomatic contours.

RESULTS

The median computational time of the CIP segmentation was 10 s. The median CIP and manually segmented volumes were 477 ml and 309 ml, respectively. CIP segmentations were significantly more robust than manual segmentations (median δCIP = 14ml, median dsiCIP = 99% vs. median δmanual = 222ml, median dsimanual = 82%) with pWilcoxon~10-16. The agreement between CIP and manual segmentations had a median DSIAgree of 60%. While 13% (47/354) of the nodules did not require any manual adjustment, minor to substantial manual adjustments were needed for 87% (305/354) of the nodules. CIP segmentations were observed to perform poorly (median DSIAgree≈50%) for non-/sub-solid nodules with subtle appearances and poorly defined boundaries.

CONCLUSION

Semi-automatic CIP segmentation can potentially reduce the physician workload for 13% of nodules owing to its computational efficiency and superior stability compared to manual segmentation. Although manual adjustment is needed for many cases, CIP segmentation provides a preliminary contour for physicians as a starting point.

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.

Automated detection and quantitative assessment of pulmonary airways depicted on CT images

We developed and tested a fully automated computerized scheme that identifies pulmonary airway sections depicted on computed tomography (CT) images and computes their sizes including the lumen and airway wall areas. The scheme includes four processing modules that (1) segment left and right lung areas, (2) identify airway locations, (3) segment airway walls from neighboring pixels, and (4) compute airway sizes. The scheme uses both a raster scanning and a labeling algorithm complemented by simple classification rules for region size and circularity to automatically search for and identify airway sections of interest. A profile tracking method is used to segment airway walls from neighboring pixels including those associated with dense tissue (i.e., pulmonary arteries) along scanning radial rays. A partial pixel membership method is used to compute airway size. The scheme was tested on ten randomly selected CT studies that included 26 sets of CT images acquired using both low and conventional dose CT examinations with one of four reconstruction algorithms (namely, "bone," "lung," "soft," and "standard" convolution kernels). Three image section thicknesses (1.25, 2.5, and 5 mm) were evaluated. The scheme detected a large number of quantifiable airway sections when the CT images were reconstructed using high spatial frequency convolution kernels. The detection results demonstrated a consistent trend for all test image sets in that as airway lumen size increases, on average the airway wall area increases as well and the wall area percentage decreases. The study suggested that CT images reconstructed using high spatial frequency convolution kernels and thin-section thickness were most amenable to automated detection, reasonable segmentation, and quantified assessment when the airways are close to being perpendicular to the CT image plane.

Pictorial essay of radiological features of benign intrathoracic masses

With increased exposure of patients to routine imaging, incidental benign intrathoracic masses are frequently recognized. Most have classical imaging features, which are pathognomonic for their benignity. The aim of this pictorial review is to educate the reader of radiological features of several types of intrathoracic masses. The masses are categorized based on their location/origin and are grouped into parenchymal, pleural, mediastinal, or bronchial. Thoracic wall masses that invade the thorax such as neurofibromas and lipomas are included as they may mimic intrathoracic masses. All examples are illustrated and include pulmonary hamartoma, pleural fibroma, sarcoidosis, bronchial carcinoid, and bronchoceles together with a variety of mediastinal cysts on plain radiographs, computed tomography (CT) and magnetic resonance imaging (MRI). Sometimes a multimodality approach would be needed to confirm the diagnosis in atypical cases. The study would include the incorporation of radionuclide studies and relevant discussion in a multidisciplinary setting.

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.

Differentiating between Subsolid and Solid Pulmonary Nodules at CT: Inter- and Intraobserver Agreement between Experienced Thoracic Radiologists

PURPOSE

To quantify the reproducibility and accuracy of experienced thoracic radiologists in differentiating between subsolid and solid pulmonary nodules at CT.

MATERIALS AND METHODS

The institutional review board of Beth Israel Deaconess Medical Center approved this multicenter study. Six thoracic radiologists, with a mean of 21 years of experience in thoracic radiology (range, 17-22 years), selected images of 10 solid and 10 subsolid nodules to create a database of 120 nodules; this selection served as the reference standard. Each radiologist then interpreted 120 randomly ordered nodules in two different sessions that were separated by a minimum of 3 weeks. The radiologists classified whether or not each nodule was subsolid. Inter- and intraobserver agreement was assessed with a κ statistic. The number of correct classifications was calculated and correlated with nodule size by using Bland-Altman plots. The relationship between disagreement and nodule morphologic characteristics was analyzed by calculating the intraclass correlation coefficient.

RESULTS

Interobserver agreement (κ) was 0.619 (range, 0.469-0.745; 95% confidence interval (CI): 0.576, 0.663) and 0.670 (range, 0.440-0.839; 95% CI: 0.608, 0.733) for interpretation sessions 1 and 2, respectively. Intraobserver agreement (κ) was 0.792 (95% CI: 0.750, 0.833). Averaged for interpretation sessions, correct classification was achieved by all radiologists for 58% (70 of 120) of nodules. Radiologists agreed with their initial determination (the reference standard) in 77% of cases (range, 45%-100%). Nodule size weakly correlated with correct classification (long axis: Spearman rank correlation coefficient, rs = 0.161 and P = .049; short axis: rs = 0.128 and P = .163).

CONCLUSION

The reproducibility and accuracy of thoracic radiologists in classifying whether or not a nodule is subsolid varied in the retrospective study. This inconsistency may affect surveillance recommendations and prognostic determinations.

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.

The assessment of the role of baseline low-dose CT scan in patients at high risk of lung cancer

BACKGROUND

Despite the progress in contemporary medicine comprising diagnostic and therapeutic methods, lung cancer is still one of the biggest health concerns in many countries of the world. The main purpose of the study was to evaluate the detection rate of pulmonary nodules and lung cancer in the initial, helical low-dose CT of the chest as well as the analysis of the relationship between the size and the histopathological character of the detected nodules.

MATERIAL/METHODS

We retrospectively evaluated 1999 initial, consecutive results of the CT examinations performed within the framework of early lung cancer detection program initiated in Szczecin. The project enrolled persons of both sexes, aged 55-65 years, with at least 20 pack-years of cigarette smoking or current smokers. The analysis included assessment of the number of positive results and the evaluation of the detected nodules in relationship to their size. All of the nodules were classified into I of VI groups and subsequently compared with histopathological type of the neoplastic and nonneoplastic pulmonary lesions.

RESULTS

Pulmonary nodules were detected in 921 (46%) subjects. What is more, malignant lesions as well as lung cancer were significantly, more frequently discovered in the group of asymptomatic nodules of the largest dimension exceeding 15 mm.

CONCLUSIONS

The initial, low-dose helical CT of the lungs performed in high risk individuals enables detection of appreciable number of indeterminate pulmonary nodules. In most of the asymptomatic patients with histopathologically proven pulmonary nodules greater than 15 mm, the mentioned lesions are malignant, what warrants further, intensified diagnostics.

The Lung Reporting and Data System (LU-RADS): A Proposal for Computed Tomography Screening

Despite the positive outcome of the recent randomized trial of computed tomography (CT) screening for lung cancer, substantial implementation challenges remain, including the clear reporting of relative risk and suggested workup of screen-detected nodules. Based on current literature, we propose a 6-level Lung-Reporting and Data System (LU-RADS) that classifies screening CTs by the nodule with the highest malignancy risk. As the LU-RADS level increases, the risk of malignancy increases. The LU-RADS level is linked directly to suggested follow-up pathways. Compared with current narrative reporting, this structure should improve communication with patients and clinicians, and provide a data collection framework to facilitate screening program evaluation and radiologist training. In overview, category 1 includes CTs with no nodules and returns the subject to routine screening. Category 2 scans harbor minimal risk, including <5 mm, perifissural, or long-term stable nodules that require no further workup before the next routine screening CT. Category 3 scans contain indeterminate nodules and require CT follow up with the interval dependent on nodule size (small [5-9 mm] or large [≥10 mm] and possibly transient). Category 4 scans are suspicious and are subdivided into 4A, low risk of malignancy; 4B, likely low-grade adenocarcinoma; and 4C, likely malignant. The 4B and 4C nodules have a high likelihood of neoplasm simply based on screening CT features, even if positron emission tomography, needle biopsy, and/or bronchoscopy are negative. Category 5 nodules demonstrate frankly malignant behavior on screening CT, and category 6 scans contain tissue-proven malignancies.

Density features of screened lung tumors in low-dose computed tomography

RATIONALE AND OBJECTIVES

Using low-dose computed tomography (LDCT), small and heterogeneous lung tumors are detected in screening. The criteria for assessing detected tumors are crucial for determining follow-up or resection strategies. The purpose of this study was to investigate the capacity of density features in differentiating lung tumors.

MATERIALS AND METHODS

From July 2008 to December 2011, 48 surgically confirmed tumors (29 malignancies, comprising 17 cases of adenocarcinoma and 12 cases of adenocarcinoma in situ [AdIs], and 19 benignancies, comprising 11 cases of atypical adenomatous hyperplasia [AAH] and eight cases of benign non-AAH) in 38 patients were retrospectively evaluated, indicating that the positive predictive value (PPV) of physicians is 60.4% (29/48). Three types of density features, tumor disappearance rate (TDR), mean, and entropy, were obtained from the CT values of detected tumors.

RESULTS

Entropy is capable of differentiating malignancy from benignancy but is limited in differentiating AdIs from benign non-AAH. The combination of entropy and TDR is effective for predicting malignancy with an accuracy of 87.5% (42/48) and a PPV of 89.7% (26/29), improving the PPV of physicians by 29.3%. The combination of entropy and mean adequately clarifies the four pathology groups with an accuracy of 72.9% (35/48). For tumors with a mean below -400 Hounsfield units, the criterion of an entropy larger than 5.4 might be appropriate for diagnosing malignancy. For others, the pathology is either benign non-AAH or adenocarcinoma; adenocarcinoma has a higher entropy than benign non-AAH, with the exception of tuberculoma.

CONCLUSIONS

Combining density features enables differentiating heterogeneous lung tumors in LDCT.

Noninvasive characterization of the histopathologic features of pulmonary nodules of the lung adenocarcinoma spectrum using computer-aided nodule assessment and risk yield (CANARY)--a pilot study

INTRODUCTION

Pulmonary nodules of the adenocarcinoma spectrum are characterized by distinctive morphological and radiologic features and variable prognosis. Noninvasive high-resolution computed tomography-based risk stratification tools are needed to individualize their management.

METHODS

Radiologic measurements of histopathologic tissue invasion were developed in a training set of 54 pulmonary nodules of the adenocarcinoma spectrum and validated in 86 consecutively resected nodules. Nodules were isolated and characterized by computer-aided analysis, and data were analyzed by Spearman correlation, sensitivity, and specificity and the positive and negative predictive values.

RESULTS

Computer-aided nodule assessment and risk yield (CANARY) can noninvasively characterize pulmonary nodules of the adenocarcinoma spectrum. Unsupervised clustering analysis of high-resolution computed tomography data identified nine unique exemplars representing the basic radiologic building blocks of these lesions. The exemplar distribution within each nodule correlated well with the proportion of histologic tissue invasion, Spearman R = 0.87, p < 0.0001 and 0.89 and p < 0.0001 for the training and the validation set, respectively. Clustering of the exemplars in three-dimensional space corresponding to tissue invasion and lepidic growth was used to develop a CANARY decision algorithm that successfully categorized these pulmonary nodules as "aggressive" (invasive adenocarcinoma) or "indolent" (adenocarcinoma in situ and minimally invasive adenocarcinoma). Sensitivity, specificity, positive predictive value, and negative predictive value of this approach for the detection of aggressive lesions were 95.4, 96.8, 95.4, and 96.8%, respectively, in the training set and 98.7, 63.6, 94.9, and 87.5%, respectively, in the validation set.

CONCLUSION

CANARY represents a promising tool to noninvasively risk stratify pulmonary nodules of the adenocarcinoma spectrum.

National lung screening trial: variability in nodule detection rates in chest CT studies

PURPOSE

To characterize the variability in radiologists' interpretations of computed tomography (CT) studies in the National Lung Screening Trial (NLST) (including assessment of false-positive rates [FPRs] and sensitivity), to examine factors that contribute to variability, and to evaluate trade-offs between FPRs and sensitivity among different groups of radiologists.

MATERIALS AND METHODS

The HIPAA-compliant NLST was approved by the institutional review board at each screening center; all participants provided informed consent. NLST radiologists reported overall screening results, nodule-specific findings, and recommendations for diagnostic follow-up. A noncalcified nodule of 4 mm or larger constituted a positive screening result. The FPR was defined as the rate of positive screening examinations in participants without a cancer diagnosis within 1 year. Descriptive analyses and mixed-effects models were utilized. The average odds ratio (OR) for a false-positive result across all pairs of radiologists was used as a measure of variability.

RESULTS

One hundred twelve radiologists at 32 screening centers each interpreted 100 or more NLST CT studies, interpreting 72 160 of 75 126 total NLST CT studies in aggregate. The mean FPR for radiologists was 28.7% ± 13.7 (standard deviation), with a range of 3.8%-69.0%. The model yielded an average OR of 2.49 across all pairs of radiologists and an OR of 1.83 for pairs within the same screening center. Mean FPRs were similar for academic versus nonacademic centers (27.9% and 26.7%, respectively) and for centers inside (25.0%) versus outside (28.7%) the U.S. "histoplasmosis belt." Aggregate sensitivity was 96.5% for radiologists with FPRs higher than the median (27.1%), compared with 91.9% for those with FPRs lower than the median (P = .02).

CONCLUSION

There was substantial variability in radiologists' FPRs. Higher FPRs were associated with modestly higher sensitivity.

Long-term surveillance of ground-glass nodules: evidence from the MILD trial

INTRODUCTION

The purpose of this study was to evaluate the natural evolution of ground-glass nodules (GGNs) in the Multicentric Italian Lung Detection (MILD) trial, which adopted a nonsurgical approach to this subset of lesions.

METHODS

From September 2005 to August 2007, 56 consecutive MILD participants with 76 GGNs were identified from 1866 individuals who underwent baseline low-dose computed tomography. The features of GGNs were assessed and compared with the corresponding repeat low-dose computed tomographies after a mean time of 50.26 ± 7.3 months. The GGNs were classified as pure (pGGN) or part-solid (psGGN) GGNs. The average of the maximum and the minimum diameters for both pGGNs and psGGNs and the maximum diameter of the solid portion of psGGNs were manually measured. At follow-up, GGNs were classified as follows: resolved, decreased, stable, or progressed (according to three defined growth patterns).

RESULTS

A total of 15 of 48 pGGNs (31.3%) resolved, 4 of 48 (8.3%) decreased in size, 21 of 48 (43.8%) remained stable, and 8 of 48 (16.7%) progressed. Among the psGGNs with a solid component smaller than 5 mm, 3 of 26 (11.5%) resolved, 11 of 26 (42.3%) remained stable, and 12 of 26 (46.2%) progressed. One of the two psGGNs with a solid component larger than 5 mm remained stable, and the other decreased in size. Four lung cancers were detected among the GGN subjects, but only one arose from a psGGN, and was resected in stage Ia.

CONCLUSIONS

The progression rate of the GGNs toward clinically relevant disease was extremely low in the MILD trial and supports an active surveillance attitude.

Computer-aided nodule detection and volumetry to reduce variability between radiologists in the interpretation of lung nodules at low-dose screening computed tomography

OBJECTIVE

The aim of this study was to evaluate whether a computer-aided diagnosis (CAD) system improves interobserver agreement in the interpretation of lung nodules at low-dose computed tomography (CT) screening for lung cancer.

MATERIALS AND METHODS

Baseline low-dose screening CT examinations from 134 participants enrolled in the National Lung Screening Trial were reviewed by 7 chest radiologists. All participants consented to the use of their deidentified images for research purposes. Screening results were classified as positive when noncalcified nodules larger than 4 mm in diameter were present. Follow-up evaluation was recommended according to the nodule diameter: 4 mm or smaller, more than 4 to 8 mm, and larger than 8 mm. When multiple nodules were present, recommendations were based on the largest nodule. Readers initially assessed the nodule presence visually and measured the average nodule diameter manually. Revision of their decisions after reviewing the CAD marks and size measurement was allowed. Interobserver agreement evaluated using multirater κ statistics was compared between initial assessment and that with CAD.

RESULTS

Multirater κ values for the positivity of the screening results and follow-up recommendations were improved from moderate (κ = 0.53-0.54) at initial assessment to good (κ = 0.66-0.67) after reviewing CAD results. The average percentage of agreement between reader pairs on the positivity of screening results and follow-up recommendations per case was also increased from 77% and 72% at initial assessment to 84% and 80% with CAD, respectively.

CONCLUSION

Computer-aided diagnosis may improve the reader agreement on the positivity of screening results and follow-up recommendations in the assessment of low-dose screening CT.

A comparison of computed tomography, computed radiography, and film-screen radiography for the detection of canine pulmonary nodules

Computed tomography (CT) has become more widely available and computed radiography (CR) has replaced film-screen radiography for canine thoracic imaging in many veterinary practices. There are limited data comparing these modalities in a veterinary clinical setting to detect pulmonary nodules. We compared CT, CR, and film-screen radiography for detecting the presence, number, and characteristics of pulmonary nodules in dogs. Observer performance for a variety of experience levels was also evaluated. Twenty-one client-owned dogs with a primary neoplastic process underwent CT and CR; nine also received film-screen radiographs. Positive/negative classification by consensus agreed between the three modalities in 8/9 dogs and between CR and CT in the remaining 12. CT detected the greatest (P = 0.002) total number of nodules and no difference was seen between CR and films. The greatest number of nodules was seen in the right middle and both caudal regions, but only using CT (P < 0.0001). Significantly smaller nodules were detected with CT (P = 0.0007) and no difference in minimum size was detected between CR and films. Observer accuracy was high for all modalities; particularly for CT (90.5-100%) and for the senior radiologist (90.5-100%). CT was also characterized by the least interobserver variability. Although CT, CR, and film-screen performed similarly in determining the presence or absence of pulmonary nodules, a greater number of smaller nodules was detected with CT, and CT was associated with greater diagnostic confidence and observer accuracy and agreement.

No benefit for consensus double reading at baseline screening for lung cancer with the use of semiautomated volumetry software

PURPOSE

To retrospectively evaluate the performance of consensus double reading compared with single reading at baseline screening of a lung cancer computed tomography (CT) screening trial.

MATERIALS AND METHODS

The study was approved by the Dutch Minister of Health and ethical committees. Written informed consent was obtained from all participants. The benefit of consensus double reading was expressed by the percentage change in cancer detection rate, recall rate, number of additional nodules detected, and change in sensitivity and specificity in 7557 participants. The reference standard was a retrospective analysis of the serial CT scans performed in participants diagnosed with lung cancer during a 2-year period after baseline. Semiautomated volumetric software was used for nodule evaluation. McNemar tests were performed to test statistical significance. In addition, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated and 95% confidence intervals (CIs) constructed.

RESULTS

Seventy-four cases of lung cancer were qualified as detectable at baseline. Compared with single reading, consensus double reading did not increase the cancer detection rate (2.7%; 95% CI: -1.0%, 6.4%; P = .50) or change the recall rate (20.6% vs 20.8%, P = .28), but led to the detection of 19.0% (1635 of 8623; 95% CI: 18.0%, 19.9%, P < .01) more nodules. The sensitivity, specificity, PPV, and NPV were 95.9% (71 of 74), 80.2% (6001 of 7483), 4.6% (71 of 1553) and 99.9% (6001 of 6004) for single reading and 98.6% (73 of 74), 80.0% (1497 of 7483), 4.6% (73 of 1570), and 99.9% (5986 of 5987) for consensus double reading, respectively.

CONCLUSION

There is no statistically significant benefit for consensus double reading at baseline screening for lung cancer with the use of a nodule management strategy based solely on semiautomated volumetry.

Enabling comparative effectiveness research with informatics: show me the data!

RATIONALE

Both outcomes researchers and informaticians are concerned with information and data. As such, some of the central challenges to conducting successful comparative effectiveness research can be addressed with informatics solutions.

METHODS

Specific informatics solutions which address how data in comparative effectiveness research are enriched, stored, shared, and analyzed are reviewed.

RESULTS

Imaging data can be made more quantitative, uniform, and structured for researchers through the use of lexicons and structured reporting. Secure and scalable storage of research data is enabled through data warehouses and cloud services. There are a number of national efforts to help researchers share research data and analysis tools.

CONCLUSION

There is a diverse arsenal of informatics tools designed to meet the needs of comparative effective researchers.

Interobserver variability in chest CT and whole body FDG-PET screening for distant metastases in head and neck cancer patients

Purpose

The aim of the study was to assess the interobserver variability in chest computed tomography (CT) and whole body 2-deoxy-2-[¹⁸F]fluoro-d-glucose positron emission tomography (FDG-PET) screening for distant metastases in head and neck squamous cell carcinoma (HNSCC) patients.

Procedure

Chest CT and whole body FDG-PET of 69 HNSCC patients with high-risk factors who underwent screening for distant metastases were analyzed. All scans were independently read by two experienced radiologists or nuclear physicians who were blinded to the other examinations and follow-up results.

Results

A kappa of 0.516 was found for assessment of size on CT. Kappa values for origin and susceptibility of 0.406 and 0.512 for CT and 0.834 and 0.939 for PET were found, respectively. The overall conclusions had a kappa of 0.517–0.634 for CT and 0.820–1.000 for PET.

Conclusions

In screening for distant metastases in HNSCC patients with high-risk factors, chest CT readings had a reasonable to substantial agreement, while PET readings showed an almost perfect agreement. These findings suggest that for optimal assessment in clinical practice, PET most often can be scored by one observer, but CT should probably more often be scored by different observers in consensus or combined with PET.

Separation of left and right lungs using 3-dimensional information of sequential computed tomography images and a guided dynamic programming algorithm

OBJECTIVE

This article presents a new computerized scheme that aims to accurately and robustly separate left and right lungs on computed tomography (CT) examinations.

METHODS

We developed and tested a method to separate the left and right lungs using sequential CT information and a guided dynamic programming algorithm using adaptively and automatically selected start point and end point with especially severe and multiple connections.

RESULTS

The scheme successfully identified and separated all 827 connections on the total 4034 CT images in an independent testing data set of CT examinations. The proposed scheme separated multiple connections regardless of their locations, and the guided dynamic programming algorithm reduced the computation time to approximately 4.6% in comparison with the traditional dynamic programming and avoided the permeation of the separation boundary into normal lung tissue.

CONCLUSIONS

The proposed method is able to robustly and accurately disconnect all connections between left and right lungs, and the guided dynamic programming algorithm is able to remove redundant processing.

Evaluation of reader variability in the interpretation of follow-up CT scans at lung cancer screening

PURPOSE

To measure reader agreement in determining whether lung nodules detected at baseline screening computed tomography (CT) had changed at subsequent screening examinations and to evaluate the variability in recommendations for further follow-up.

MATERIALS AND METHODS

All subjects were enrolled in the National Lung Screening Trial (NLST), and each participant consented to the use of their de-identified images for research purposes. The authors randomly selected 100 cases of nodules measuring at least 4.0 mm at 1-year screening CT that were considered by the original screening CT reader to be present on baseline CT scans; nodules considered by the original reader to have changed were oversampled. Selected images from each case showing the entire nodule at both examinations were preloaded on a picture archiving and communication system workstation. Nine radiologists served as readers, and they evaluated whether the nodule was present at baseline and recorded the bidimensional measurements and nodule characteristics at each examination, presence or absence of change, results of screening CT, and follow-up recommendations (high-level follow-up, low-level follow-up, no follow-up).

RESULTS

On the basis of reviews during case selection, five nodules seen at follow-up were judged not to have been present at baseline; for 19 of the remaining 95 cases, at least one reader judged the nodule not to have been present at baseline. For the 76 nodules that were unanimously considered to have been present at baseline, 21%-47% (mean ± standard deviation, 30% ± 9) were judged to have grown. The κ values were similar for growth (κ = 0.55) and a positive screening result (κ = 0.51) and were lower for a change in margins and attenuation (κ = 0.27-0.31). The κ value in the recommendation of high- versus low-level follow-up was high (κ = 0.66).

CONCLUSION

Reader agreement on nodule growth and screening result was moderate to substantial. Agreement on follow-up recommendations was lower.

Computer-aided detection of early interstitial lung diseases using low-dose CT images

This study aims to develop a new computer-aided detection (CAD) scheme to detect early interstitial lung disease (ILD) using low-dose computed tomography (CT) examinations. The CAD scheme classifies each pixel depicted on the segmented lung areas into positive or negative groups for ILD using a mesh-grid-based region growth method and a multi-feature-based artificial neural network (ANN). A genetic algorithm was applied to select optimal image features and the ANN structure. In testing each CT examination, only pixels selected by the mesh-grid region growth method were analyzed and classified by the ANN to improve computational efficiency. All unselected pixels were classified as negative for ILD. After classifying all pixels into the positive and negative groups, CAD computed a detection score based on the ratio of the number of positive pixels to all pixels in the segmented lung areas, which indicates the likelihood of the test case being positive for ILD. When applying to an independent testing dataset of 15 positive and 15 negative cases, the CAD scheme yielded the area under receiver operating characteristic curve (AUC = 0.884 ± 0.064) and 80.0% sensitivity at 85.7% specificity. The results demonstrated the feasibility of applying the CAD scheme to automatically detect early ILD using low-dose CT examinations.

Computerized comprehensive data analysis of lung imaging database consortium (LIDC)

PURPOSE

Lung Image Database Consortium (LIDC) is the largest public CT image database of lung nodules. In this study, the authors present a comprehensive and the most updated analysis of this dynamically growing database under the help of a computerized tool, aiming to assist researchers to optimally use this database for lung cancer related investigations.

METHODS

The authors developed a computer scheme to automatically match the nodule outlines marked manually by radiologists on CT images. A large variety of characteristics regarding the annotated nodules in the database including volume, spiculation level, elongation, interobserver variability, as well as the intersection of delineated nodule voxels and overlapping ratio between the same nodules marked by different radiologists are automatically calculated and summarized. The scheme was applied to analyze all 157 examinations with complete annotation data currently available in LIDC dataset.

RESULTS

The scheme summarizes the statistical distributions of the abovementioned geometric and diagnosis features. Among the 391 nodules, (1) 365 (93.35%) have principal axis length < or =20 mm; (2) 120, 75, 76, and 120 were marked by one, two, three, and four radiologists, respectively; and (3) 122 (32.48%) have the maximum volume overlapping ratios -80% for the delineations of two radiologists, while 198 (50.64%) have the maximum volume overlapping ratios <60%. The results also showed that 72.89% of the nodules were assessed with malignancy score between 2 and 4, and only 7.93% of these nodules were considered as severely malignant (malignancy > or =4).

CONCLUSIONS

This study demonstrates that LIDC contains examinations covering a diverse distribution of nodule characteristics and it can be a useful resource to assess the performance of the nodule detection and/or segmentation schemes.

Extracardiac findings in cardiac computed tomographic angiography in patients at low to intermediate risk for coronary artery disease

PURPOSE

To evaluate the prevalence, clinical significance, interobserver agreement, and follow-up of extracardiac findings on coronary computed tomographic angiography (CTA).

METHODS

A prospectively recruited cohort of 80 patients at low to intermediate risk of having coronary artery disease underwent CTA with field of view imaging from lung apices to upper abdomen. Two staff radiologists read each scan independently. Scans read by reader no. 1 were read as part of routine clinical practice, and the findings were subsequently reclassified to potentially significant, as defined by requiring clinical or radiologic follow-up, and insignificant by a separate observer, whereas reader no. 2 retrospectively read and autonomously classified the findings as potentially significant or insignificant.

RESULTS

Reader no. 1 found 7 potentially significant findings in 7 patients and 33 insignificant findings in 29 patients. Reader no. 2 found 10 potentially significant findings in 10 patients and 59 insignificant findings in 42 patients. Inter-rater agreement was moderate (kappa = 0.49; 95% confidence interval, 0.31-0.67) for the presence vs the absence of extracardiac findings and moderate (kappa = 0.52; 95% confidence interval, 0.15-0.89) for the presence of potentially significant extracardiac findings. The most common potentially significant finding was possibly malignant lung nodule (n = 6 [reader 1], 4 [reader 2]). Four patients with potentially significant findings received follow-up imaging, and 1 patient underwent biopsy, which was complicated by pneumothorax. No diagnoses of malignancy were made.

CONCLUSIONS

Extracardiac findings are frequent and moderately reproducible, however, in this study, not associated with clinical benefit. Large prospective studies are required to establish whether reporting of extracardiac findings is associated with improved patient outcomes.

Observer accuracy in the detection of pulmonary nodules on CT: effect of cine frame rate

AIM

To assess the effect of cine frame rate on the accuracy of the detection of pulmonary nodules at computed tomography (CT).

MATERIALS AND METHODS

CT images of 15 consecutive patients with (n = 13) or without (n = 2) pulmonary metastases were identified. Initial assessment by two thoracic radiologists provided the "actual" or reference reading. Subsequently, 10 radiologists [board certified radiologists (n = 4) or radiology residents (n = 6)] used different fixed cine frame rates for nodule detection. Within-subjects analysis of variance (ANOVA) was used to evaluate the data.

RESULTS

Eighty-nine nodules were identified by the thoracic radiologists (median 8, range 0-29 per patient; median diameter 9 mm, range 4-40 mm). There was a non-statistically significant trend to reduced accuracy at higher frame rates (p=0.113) with no statistically significant difference between experienced observers and residents (p = 0.79).

CONCLUSION

The accuracy of pulmonary nodule detection at higher cine frame rates is reduced, unrelated to observer experience.

Pleural nodule identification in low-dose and thin-slice lung computed tomography

A completely automated system for the identification of pleural nodules in low-dose and thin-slice computed tomography (CT) of the lung has been developed. The directional-gradient concentration method has been applied to the pleura surface and combined with a morphological opening-based procedure to generate a list of nodule candidates. Each nodule candidate is characterized by 12 morphological and textural features, which are analyzed by a rule-based filter and a neural classifier. This detection system has been developed and validated on a dataset of 42 annotated CT scans. The k-fold cross validation has been used to evaluate the neural classifier performance. The system performance variability due to different ground truth agreement levels is discussed. In particular, the poor 44% sensitivity obtained on the ground truth with agreement level 1 (nodules annotated by only one radiologist) with six FP per scan grows up to the 72% if the underlying ground truth is changed to the agreement level 2 (nodules annotated by two radiologists).

Assessment of radiologist performance in the detection of lung nodules: dependence on the definition of "truth"

RATIONALE AND OBJECTIVES

Studies that evaluate the lung nodule detection performance of radiologists or computerized methods depend on an initial inventory of the nodules within the thoracic images (the "truth"). The purpose of this study was to analyze (1) variability in the "truth" defined by different combinations of experienced thoracic radiologists and (2) variability in the performance of other experienced thoracic radiologists based on these definitions of "truth" in the context of lung nodule detection in computed tomographic (CT) scans.

MATERIALS AND METHODS

Twenty-five thoracic CT scans were reviewed by four thoracic radiologists, who independently marked lesions they considered to be nodules >or=3 mm in maximum diameter. Panel "truth" sets of nodules were then derived from the nodules marked by different combinations of two and three of these four radiologists. The nodule detection performance of the other radiologists was evaluated based on these panel "truth" sets.

RESULTS

The number of "true" nodules in the different panel "truth" sets ranged from 15 to 89 (mean 49.8 +/- 25.6). The mean radiologist nodule detection sensitivities across radiologists and panel "truth" sets for different panel "truth" conditions ranged from 51.0 to 83.2%; mean false-positive rates ranged from 0.33 to 1.39 per case.

CONCLUSIONS

Substantial variability exists across radiologists in the task of lung nodule identification in CT scans. The definition of "truth" on which lung nodule detection studies are based must be carefully considered, because even experienced thoracic radiologists may not perform well when measured against the "truth" established by other experienced thoracic radiologists.

Performance evaluation of a computer-aided detection algorithm for solid pulmonary nodules in low-dose and standard-dose MDCT chest examinations and its influence on radiologists

The aim of the study was to evaluate the performance of a computer-aided detection (CAD) algorithm in low-dose and full-dose multidetector-row CT (MDCT) of the thorax and its impact on radiologists' performance. Chest CT examinations of 77 patients were evaluated retrospectively for pulmonary nodules. All patients had undergone a 16-slice MDCT chest examination with a standard acquisition protocol. Artificial image noise was added to the raw data to simulate image acquisition at 10 mAs(eff.) The data were transferred to dedicated lung analysis software (LungCare) with a prototype CAD algorithm (LungCAD). CAD was applied to both dose settings. Images were read by a radiologist and a first-year resident with and without the software at both dose settings. All images were reviewed in consensus by the two radiologists to set the reference standard. Sensitivity results with respect to the reference standard were compared. No statistically significant differences in the detection rate for all pulmonary nodules could be found between low-dose and full-dose settings for the CAD software alone (p = 0.0065). Both radiologists displayed a statistically significant increase in sensitivity with the use of CAD (p<0.0001). In conclusion, CAD is beneficial in both low-dose and standard-dose settings. This may be beneficial in reducing false-negative diagnosis in lung cancer screening, standard chest examinations and the search for metastases.

An automated CT based lung nodule detection scheme using geometric analysis of signed distance field

The authors present a new computerized scheme to automatically detect lung nodules depicted on computed tomography (CT) images. The procedure is performed in the signed distance field of the CT images. To obtain an accurate signed distance field, CT images are first interpolated linearly along the axial direction to form an isotropic data set. Then a lung segmentation strategy is applied to smooth the lung border aiming to include as many juxtapleural nodules as possible while minimizing over segmentations of the lung regions. Potential nodule regions are then detected by locating local maximas of signed distances in each subvolume with values and the sizes larger than the smallest nodule of interest in the three-dimensional space. Finally, all detected candidates are scored by computing the similarity distance of their medial axis-like shapes obtained through a progressive clustering strategy combined with a marching cube algorithm from a sphere based shape. A free-response receiver operating characteristics curve is computed to assess the scheme performance. A performance test on 52 low-dose CT screening examinations that depict 184 verified lung nodules showed that during the initial stage the scheme achieved an asymptotic maximum sensitivity of 95.1% (175/184) with an average of 1200 suspicious voxels per CT examination. The nine missed nodules included two small solid nodules (with a diameter < or =3.1 mm) and seven nonsolid nodules. The final performance level after the similarity scoring stage was an absolute sensitivity level, namely, including the nine missed during the initial stage, of 81.5% (150/184) with 6.5 false-positive identifications per CT examination. This preliminary study demonstrates the feasibility of applying a simple and robust geometric model using the signed distance field to identify suspicious lung nodules. In the authors' data set the sensitivity of this scheme is not affected by nodule size. In addition to potentially being a stand alone approach, the signed distance field based method can be easily implemented as an initial filtering step in other computer-aided detection schemes.

[Usefulness of maximum intensity projections in low-radiation multislice CT lung cancer screening]

OBJECTIVE

To evaluate the diagnostic accuracy of non-overlapping 10-mm-thick axial maximum intensity projections (MIP) in the detection of pulmonary nodules in subjects participating in a lung cancer screening program (LCSP) using multislice computed tomography (MSCT) with a low dose of radiation.

MATERIAL AND METHODS

We evaluated 52 consecutive low-radiation MSCT studies in asymptomatic smokers included in an LCSP (1.25 mm axial images). Axial MIPs with 10mm slice thickness (30 images) were performed and evaluated retrospectively; readers were blind to the initial radiological report. All nodules detected were considered, regardless of their size or consistency. The standard of reference was determined by double reading and consensus for each nodule.

RESULTS

A total of 162 pulmonary nodules (mean size: 3.9 mm, sd: 1.7) were detected. MIP reconstruction detected 150 nodules (S = 92.6%). The initial radiological evaluation detected 108 nodules (S = 66.7%). MIP reconstruction detected 54 (33.3%) nodules that were not reported initially (mean size: 3.4 mm; sd: 1.2) but failed to detect 12 (7.4%) of the nodules reported initially (mean size: 2.91 mm; sd: 0.8). MIP detected all 35 nodules > or = 5 mm, (S =100), whereas the initial radiological evaluation only detected 27 (S = 77%). MIP reconstruction enabled more of the nodules to be detected than the 1.25-mm conventional axial slices (p < 0.01).

CONCLUSION

The introduction of non-overlapping 10-mm-thick axial MIP reconstructions in a low-radiation LCSP using MSCT enabled nodules more accurate and faster detection of pulmonary nodules in comparison with 1.25 mm conventional axial slices.

Indolent lung cancers--time for a paradigm shift: a review

The present paradigm is that slow-growing and clinically unimportant lung cancers are very rare. Over the years, convincing evidence for their existence in significant numbers has been dismissed, but with computerized tomography scanning being increasingly common, their reality can no longer be denied. The time is now ripe for a paradigm shift--with consequences for screening as well as clinical practice in the future. The doubling time of the tumor is probably the most important factor for survival of the patient, and research on how to measure this figure is urgent. Age, smoking habits, and comorbidities of the patient are now the main factors when discussing treatment with the patient; once we can predict the behavior of the cancer, i.e., whether it is indolent or actually threatens the patient's life, we will be able to give better advice.

The prevalence of extracardiac findings by multidetector computed tomography before atrial fibrillation ablation

BACKGROUND AND OBJECTIVES

The study was designed to determine the prevalence of extracardiac findings discovered during multidetector computed tomography (CT) (MDCT) examinations before atrial fibrillation ablation. Multidetector CT has become a valuable tool in detailing left atrial anatomy before catheter ablation. The incidence of extracardiac findings has been reported for electron beam CT calcium scoring and coronary MDCT, but no data exist for the prevalence of extracardiac findings discovered before atrial fibrillation ablation with MDCT.

METHODS AND RESULTS

Clinical reports from MDCT examinations before atrial fibrillation ablation and interpretations by 2 radiologists blinded to the clinical reports were reviewed for significant additional extracardiac findings and recommendations for follow-up. In 149 patients who underwent MDCT, the mean age was 55.9 +/- 11.0 years, 75% were men, and 47% had a history of smoking. Extracardiac findings were identified in 69% of patients with clinical, 90% of reader 1, and 97% of reader 2 interpretations (kappa = 0.086). Follow-up was recommended in 30% of clinical, 50% of reader 1, and 38% of reader 2 interpretations (kappa = 0.408). Pulmonary nodules were the most common additional finding and reason for suggested follow-up for all interpreters.

CONCLUSIONS

The prevalence of extracardiac abnormalities detected by MDCT is considerable. Significant variability in their identification exists between interpreters, but there is good agreement about the need for further follow-up. It is important that those who interpret these examinations are adequately trained in the identification and interpretation of both cardiac and extracardiac findings.

The Lung Image Database Consortium (LIDC) data collection process for nodule detection and annotation

RATIONALE AND OBJECTIVES

The Lung Image Database Consortium (LIDC) is developing a publicly available database of thoracic computed tomography (CT) scans as a medical imaging research resource to promote the development of computer-aided detection or characterization of pulmonary nodules. To obtain the best estimate of the location and spatial extent of lung nodules, expert thoracic radiologists reviewed and annotated each scan. Because a consensus panel approach was neither feasible nor desirable, a unique two-phase, multicenter data collection process was developed to allow multiple radiologists at different centers to asynchronously review and annotate each CT scan. This data collection process was also intended to capture the variability among readers.

MATERIALS AND METHODS

Four radiologists reviewed each scan using the following process. In the first or "blinded" phase, each radiologist reviewed the CT scan independently. In the second or "unblinded" review phase, results from all four blinded reviews were compiled and presented to each radiologist for a second review, allowing the radiologists to review their own annotations together with the annotations of the other radiologists. The results of each radiologist's unblinded review were compiled to form the final unblinded review. An XML-based message system was developed to communicate the results of each reading.

RESULTS

This two-phase data collection process was designed, tested, and implemented across the LIDC. More than 500 CT scans have been read and annotated using this method by four expert readers; these scans either are currently publicly available at http://ncia.nci.nih.gov or will be in the near future.

CONCLUSIONS

A unique data collection process was developed, tested, and implemented that allowed multiple readers at distributed sites to asynchronously review CT scans multiple times. This process captured the opinions of each reader regarding the location and spatial extent of lung nodules.

The Lung Image Database Consortium (LIDC): ensuring the integrity of expert-defined "truth"

RATIONALE AND OBJECTIVES

Computer-aided diagnostic (CAD) systems fundamentally require the opinions of expert human observers to establish "truth" for algorithm development, training, and testing. The integrity of this "truth," however, must be established before investigators commit to this "gold standard" as the basis for their research. The purpose of this study was to develop a quality assurance (QA) model as an integral component of the "truth" collection process concerning the location and spatial extent of lung nodules observed on computed tomography (CT) scans to be included in the Lung Image Database Consortium (LIDC) public database.

MATERIALS AND METHODS

One hundred CT scans were interpreted by four radiologists through a two-phase process. For the first of these reads (the "blinded read phase"), radiologists independently identified and annotated lesions, assigning each to one of three categories: "nodule >or=3 mm," "nodule <3 mm," or "non-nodule >or=3 mm." For the second read (the "unblinded read phase"), the same radiologists independently evaluated the same CT scans, but with all of the annotations from the previously performed blinded reads presented; each radiologist could add to, edit, or delete their own marks; change the lesion category of their own marks; or leave their marks unchanged. The post-unblinded read set of marks was grouped into discrete nodules and subjected to the QA process, which consisted of identification of potential errors introduced during the complete image annotation process and correction of those errors. Seven categories of potential error were defined; any nodule with a mark that satisfied the criterion for one of these categories was referred to the radiologist who assigned that mark for either correction or confirmation that the mark was intentional.

RESULTS

A total of 105 QA issues were identified across 45 (45.0%) of the 100 CT scans. Radiologist review resulted in modifications to 101 (96.2%) of these potential errors. Twenty-one lesions erroneously marked as lung nodules after the unblinded reads had this designation removed through the QA process.

CONCLUSIONS

The establishment of "truth" must incorporate a QA process to guarantee the integrity of the datasets that will provide the basis for the development, training, and testing of CAD systems.