Lung image database consortium: developing a resource for the medical imaging research community

Supervised probabilistic segmentation of pulmonary nodules in CT scans

An automatic method for lung nodule segmentation from computed tomography (CT) data is presented that is different from previous work in several respects. Firstly, it is supervised; it learns how to obtain a reliable segmentation from examples in a training phase. Secondly, the method provides a soft, or probabilistic segmentation, thus taking into account the uncertainty inherent in this segmentation task. The method is trained and tested on a public data set of 23 nodules for which soft labelings are available. The new method is shown to outperform a previously published conventional method. By merely changing the training data, non-solid nodules can also be segmented.

An Analysis of Early Studies Released by the Lung Imaging Database Consortium (LIDC)

Lung cancer remains an ongoing problem resulting in substantial deaths in the United States and the world. Within the United states, cancer of the lung and bronchus are the leading causes of fatal malignancy and make up 32% of the cancer deaths among men and 25% of the cancer deaths among women. Five year survival is low, (14%), but recent studies are beginning to provide some hope that we can increase survivability of lung cancer provided that the cancer is caught and treated in early stages. These results motivate revisiting the concept of lung cancer screening using thin slice multidetector computed tomography (MDCT) protocols and automated detection algorithms to facilitate early detection. In this environment, resources to aid Computer Aided Detection (CAD) researchers to rapidly develop and harden detection and diagnostic algorithms may have a significant impact on world health. The National Cancer Institute (NCI) formed the Lung Imaging Database Consortium (LIDC) to establish a resource for detecting, sizing, and characterizing lung nodules. This resource consists of multiple CT chest exams containing lung nodules that seveal radiologists manually countoured and characterized. Consensus on the location of the nodule boundaries, or even on the existence of a nodule at a particular location in the lung was not enforced, and each contour is considered a possible nodule. The researcher is encouraged to develop measures of ground truth to reconcile the multiple radiologist marks. This paper analyzes these marks to determine radiologist agreement and to apply statistical tools to the generation of a nodule ground truth. Features of the resulting consensus and individual markings are analyzed.

Lung cancer screening

PURPOSE OF REVIEW

Advances in imaging technologies and biomarker research offer hope that the incidence and mortality of lung cancer can be reduced by screening similar to what have been achieved for cancer of the cervix, breast, and colon.

RECENT FINDINGS

Spiral computed tomography with multitrack scanners and autofluorescence bronchoscopy offer unprecedented sensitivity to detect lung cancer even during the preinvasive stage. The high sensitivity of these tests, however, is associated with a low specificity. Better selection of individuals at highest risk of lung cancer using biomarkers in sputum, blood, or exhaled breath, as well as a better understanding of genetic susceptibility, may improve their positive predictive values, minimize unnecessary downstream investigations or treatment, as well as reduce screening costs.

SUMMARY

Improvement in the performance of sputum, exhaled breath, or blood biomarkers holds promise as the first screening step to identify individuals at highest risk of lung cancer beyond what age and smoking could predict to select those who would obtain the most benefits from spiral computed tomography or autofluorescence bronchoscopy as localization tools.

Value of axial and coronal maximum intensity projection (MIP) images in the detection of pulmonary nodules by multislice spiral CT: comparison with axial 1-mm and 5-mm slices

The purpose of this study was to investigate the diagnostic accuracy of non-overlapping 10-mm axial and coronal maximum intensity projections (MIP) in comparison with standard axial 1-mm and 5-mm slices in the detection of pulmonary nodules. Sixty patients with suspected nodules who underwent multislice spiral CT of the chest were evaluated. Axial 1-mm and 5-mm slices as well as non-overlapping 10-mm axial/coronal MIPs were interpreted independently by three blinded radiologists. After initial review, a retrospective consensus session was performed for agreement on final nodule counts using the axial 1-mm slices as gold standard. Small nodules of less than 5 mm in size were most accurately detected by the axial MIPs. Receiver operating characteristic (ROC) analysis of these small nodules showed that 5-mm slices were not capable of a statistically significant differentiation of nodules from other focal lesions in two observers (p=0.034 and p=0.012, respectively) whereas 1-mm slices and coronal/axial MIPs did allow a statistically significant differentiation in all observers (p<0.001). Nodules larger than 5 mm were equally well depicted with all modalities. Non-overlapping 10-mm axial MIPs improve the accuracy in the detection of small pulmonary nodules.

Lung Nodule Detection and Characterization With Multi-Slice CT

The influence of MSCT on nodule detection and characterization will be discussed. The objective is to improve understanding of the clinical issues involved in nodule detection, characterization, and management in light of technological advances. Topics to be covered are noninvasive characterization techniques, such as morphologic and density inspection on CT, nodule enhancement techniques, CT-PET, temporal nodule size assessment, and computer aided diagnosis for both detection and characterization.

Digital biology: an emerging and promising discipline

This article examines the role of computation and quantitative methods in modern biomedical research to identify emerging scientific, technical, policy and organizational trends. It identifies common concerns and practices in the emerging community of computationally-oriented bio-scientists by reviewing a national symposium, Digital Biology: the Emerging Paradigm, held at the National Institutes of Health in Bethesda, Maryland, November 6th and 7th 2003. This meeting showed how biomedical computing promises scientific breakthroughs that will yield significant health benefits. Three key areas that define the emerging discipline of digital biology are: scientific data integration, multi-scale modeling and networked science. Each area faces unique technical challenges and information policy issues that must be addressed as the field matures. Here we summarize the emergent challenges and offer suggestions to academia, industry and government on how best to expand the role of computation in their scientific activities.

Automated detection of pulmonary nodules on CT images: effect of section thickness and reconstruction interval--initial results

Institutional review board approval was obtained. Informed patient consent was not required. Study was compliant with HIPAA. Performance of an automated pulmonary nodule detection program was evaluated on multi-detector row CT images that were acquired once but reconstructed retrospectively at different section thicknesses and reconstruction intervals. From raw CT data in 10 patients with pulmonary nodules, three sets of CT images were reconstructed separately in each patient by selecting two section thickness and reconstruction combinations, respectively: thin group, 1 and 1 mm; overlap group, 5 and 1 mm; and thick group, 5 and 5 mm. Nodules 3 mm in diameter and larger were detected in each group (thin group, 126 nodules; overlap group, 121 nodules; and thick group, 114 nodules) by means of consensus of two radiologists. Findings were used as the reference standard for evaluation of the computer-aided detection (CAD) program. Sensitivity and number of false-positive findings per patient by CAD were: thin group, 95.2% (120 of 126 nodules) and 5.4 findings; overlap group, 94.2% (114 of 121 nodules) and 9.7 findings; and thick group, 88.6% (101 of 114 nodules) and 23.6 findings, indicating that nodule detection degraded with increase in section thickness but improved substantially with a small reconstruction interval.

Pulmonary nodules: automated detection on CT images with morphologic matching algorithm--preliminary results

Institutional review board approval was obtained. Informed patient consent was not required for this retrospective study, which involved review of previously obtained image data. Patient confidentiality was protected; the study was compliant with the Health Insurance Portability and Accountability Act. An automated pulmonary nodule detection program that takes advantage of three-dimensional volumetric data was developed and tested with multi-detector row computed tomographic (CT) images from 20 patients (13 men, seven women; age range, 40-75 years) with pulmonary nodules. A total of 164 nodules 3 mm in diameter and larger were detected by two radiologists in consensus and were used as a reference standard to evaluate the computer-aided detection (CAD) program. The CAD algorithm was structured to process nodules that were categorized into three types: isolated, juxtapleural, and juxtavascular. Overall sensitivity for nodule detection with the CAD program was 95.1% (156 of 164 nodules). The sensitivity according to nodule size was 91.2% (52 of 57 nodules) for nodules 3 mm to less than 5 mm and 97.2% (104 of 107 nodules) for nodules 5 mm and larger. The number of false-positive detections per patient was 6.9 for false nodule structures 3 mm and larger and 4.0 for false nodule structures 5 mm and larger.

New Developments in Lung Cancer Screening

Lung cancer is the most lethal cancer in our society. Late diagnosis of this disease is a major problem and so recent favorable reports with spiral computed tomography screening of high-risk populations have rekindled interest in improving early lung cancer detections. The process of lung cancer screening is a complicated process that involves many component activities. Interest to date has heavily focused on the initial case identification, but more recent reports have suggested that the issues with case work-up and surgical management also bear closer consideration. Given the dynamic nature of spiral computed tomography scan development and the remarkable improvements in imaging resolution over the last decade, there is an urgent need for research to establish optimal clinical management of early lung cancer detected in a screening setting.

Evaluation of automated lung nodule detection on low-dose computed tomography scans from a lung cancer screening program(1)

RATIONALE AND OBJECTIVES

The purpose of this study was to evaluate the performance of a fully automated lung nodule detection method in a large database of low-dose computed tomography (CT) scans from a lung cancer screening program. Because nodules demonstrate a spectrum of radiologic appearances, the performance of the automated method was evaluated on the basis of nodule malignancy status, size, subtlety, and radiographic opacity.

MATERIALS AND METHODS

A database of 393 thick-section (10 mm) low-dose CT scans was collected. Automated lung nodule detection proceeds in two phases: gray-level thresholding for the initial identification of nodule candidates, followed by the application of a rule-based classifier and linear discriminant analysis to distinguish between candidates that correspond to actual lung nodules and candidates that correspond to non-nodules. Free-response receiver operating characteristic analysis was used to evaluate the performance of the method based on a jackknife training/testing approach.

RESULTS

An overall nodule detection sensitivity of 70% (330 of 470) was attained with an average of 1.6 false-positive detections per section. At the same false-positive rate, 83% (57 of 69) of the malignant lung nodules in the database were detected. When the method was trained specifically for malignant nodules, a sensitivity of 80% (55 of 69) was attained with 0.85 false-positives per section.

CONCLUSION

We have evaluated an automated lung nodule detection method with a large number of low-dose CT scans from a lung cancer screening program. An overall sensitivity of 80% for malignant nodules was achieved with 0.85 false-positive detections per section. Such a computerized lung nodule detection method is expected to become an important part of CT-based lung cancer screening programs.