Bartjan de Hoop

Automatic lung segmentation from thoracic computed tomography scans using a hybrid approach with error detection

Lung segmentation is a prerequisite for automated analysis of chest CT scans. Conventional lung segmentation methods rely on large attenuation differences between lung parenchyma and surrounding tissue. These methods fail in scans where dense abnormalities are present, which often occurs in clinical data. Some methods to handle these situations have been proposed, but they are too time consuming or too specialized to be used in clinical practice. In this article, a new hybrid lung segmentation method is presented that automatically detects failures of a conventional algorithm and, when needed, resorts to a more complex algorithm, which is expected to produce better results in abnormal cases. In a large quantitative evaluation on a database of 150 scans from different sources, the hybrid method is shown to perform substantially better than a conventional approach at a relatively low increase in computational cost.

Pulmonary Ground-Glass Nodules: Increase in Mass as an Early Indicator of Growth

PURPOSE To compare manual measurements of diameter, volume, and mass of pulmonary ground-glass nodules (GGNs) to establish which method is best for identifying malignant GGNs by determining change across time. MATERIALS AND METHODS In this ethics committee-approved retrospective study, baseline and follow-up CT examinations of 52 GGNs detected in a lung cancer screening trial were included, resulting in 127 GGN data sets for evaluation. Two observers measured GGN diameter with electronic calipers, manually outlined GGNs to obtain volume and mass, and scored whether a solid component was present. Observer 1 repeated all measurements after 2 months. Coefficients of variation and limits of agreement were calculated by using Bland-Altman methods. In a subgroup of GGNs containing all resected malignant lesions, the ratio between intraobserver variability and growth (growth-to-variability ratio) was calculated for each measurement technique. In this subgroup, the mean time for growth to exceed the upper limit of agreement of each measurement technique was determined. RESULTS The kappa values for intra- and interobserver agreement for identifying a solid component were 0.55 and 0.38, respectively. Intra- and interobserver coefficients of variation were smallest for GGN mass (P < .001). Thirteen malignant GGNs were resected. Mean growth-to-variability ratios were 11, 28, and 35 for diameter, volume, and mass, respectively (P = .03); mean times required for growth to exceed the upper limit of agreement were 715, 673, and 425 days, respectively (P = .02). CONCLUSION Mass measurements can enable detection of growth of GGNs earlier and are subject to less variability than are volume or diameter measurements.

Comparing and combining algorithms for computer-aided detection of pulmonary nodules in computed tomography scans: The ANODE09 study

Numerous publications and commercial systems are available that deal with automatic detection of pulmonary nodules in thoracic computed tomography scans, but a comparative study where many systems are applied to the same data set has not yet been performed. This paper introduces ANODE09 ( http://anode09.isi.uu.nl), a database of 55 scans from a lung cancer screening program and a web-based framework for objective evaluation of nodule detection algorithms. Any team can upload results to facilitate benchmarking. The performance of six algorithms for which results are available are compared; five from academic groups and one commercially available system. A method to combine the output of multiple systems is proposed. Results show a substantial performance difference between algorithms, and demonstrate that combining the output of algorithms leads to marked performance improvements.

Prostate cancer: detection of lymph node metastases outside the routine surgical area with ferumoxtran-10-enhanced MR imaging.

PURPOSE To prospectively evaluate the feasibility of magnetic resonance (MR) imaging with ferumoxtran-10 in patients with prostate cancer to depict lymph node metastases outside the routine pelvic lymph node dissection (PLND) area. MATERIALS AND METHODS The study was approved by the institutional review boards at all four hospitals; patients provided written informed consent. Two hundred ninety-six consecutive men (mean age, 67 years; range, 47-83 years) with prostate cancer and an intermediate-to-high risk for nodal metastases (prostate-specific antigen level >10 ng/mL, Gleason score >6, or stage T3 disease) were enrolled. MR lymphography of the pelvis was performed 24 hours after intravenous drip infusion of ferumoxtran-10. Positive nodes at MR lymphography were indicated to be inside or outside the routine dissection area (RDA). On the basis of MR lymphography computed tomographic (CT)-guided biopsy, routine PLND, or MR imaging-guided minimal extended PLND was performed. RESULTS MR lymphography findings were positive in 58 patients. Of these, 44 had histopathologic confirmation of lymph node metastases. In 18 of 44 patients (41%), MR lymphography findings showed nodes exclusively outside the RDA, which were confirmed with MR lymphography-guided extended PLND (n = 13) and CT-guided biopsy (n = 5). In another 18 patients (41%), positive nodes were located both inside and outside the RDA at MR lymphography. In these 18 patients, routine PLND was used to confirm the nodes inside the RDA (n = 11); CT-guided biopsy was used to confirm nodes outside the RDA (n = 7). In the remaining eight patients, MR lymphography findings showed only nodes inside the RDA, which was confirmed with PLND (n = 5) and CT-guided biopsy (n = 3). In 14 of the 58 patients (24%), there was no histologic confirmation. CONCLUSION In 41% of patients with prostate cancer, nodal metastases outside the area of routine PLND were detected by using MR imaging with ferumoxtran-10.

Automatic classification of pulmonary peri-fissural nodules in computed tomography using an ensemble of 2D views and a convolutional neural network out-of-the-box

In this paper, we tackle the problem of automatic classification of pulmonary peri-fissural nodules (PFNs). The classification problem is formulated as a machine learning approach, where detected nodule candidates are classified as PFNs or non-PFNs. Supervised learning is used, where a classifier is trained to label the detected nodule. The classification of the nodule in 3D is formulated as an ensemble of classifiers trained to recognize PFNs based on 2D views of the nodule. In order to describe nodule morphology in 2D views, we use the output of a pre-trained convolutional neural network known as OverFeat. We compare our approach with a recently presented descriptor of pulmonary nodule morphology, namely Bag of Frequencies, and illustrate the advantages offered by the two strategies, achieving performance of AUC = 0.868, which is close to the one of human experts.

Pulmonary Perifissural Nodules on CT Scans: Rapid Growth Is Not a Predictor of Malignancy

PURPOSE To assess the prevalence, natural course, and malignancy rate of perifissural nodules (PFNs) in smokers participating in a lung cancer screening trial. MATERIALS AND METHODS As part of the ethics-committee approved Dutch-Belgian Randomised Lung Cancer Multi-Slice Screening Trial (NELSON), computed tomography (CT) was used to screen 2994 current or former heavy smokers, aged 50-74 years, for lung cancer. CT was repeated after 1 and 3 years, with additional follow-up CT scans if necessary. All baseline CT scans were screened for nodules. Nodule volume was determined with automated volumetric analysis. Homogeneous solid nodules, attached to a fissure with a lentiform or triangular shape, were classified as PFNs. Nodules were considered benign if they did not grow during the total follow-up period or were proved to be benign in a follow-up by a pulmonologist. Prevalence, growth, and malignancy rate of PFNs were assessed. RESULTS At baseline screening, 4026 nodules were detected in 1729 participants, and 19.7% (794 of 4026) of the nodules were classified as PFNs. The mean size of the PFNs was 4.4 mm (range: 2.8-10.6 mm) and the mean volume was 43 mm3 (range: 13-405 mm3). None of the PFNs were found to be malignant during follow-up. Between baseline and the first follow-up CT scan, 15.5% (123 of 794) were found to have grown, and 8.3% (66 of 794) had a volume doubling time of less than 400 days. One PFN was resected and proved to be a lymph node. CONCLUSION PFNs are frequently found at CT scans for lung cancer. They can show growth rates in the range of malignant nodules, but none of the PFNs in the present study turned out to be malignant. Recognition of PFNs can reduce the number of follow-up examinations required for the workup of suspicious nodules.

Automatic Segmentation of Pulmonary Lobes Robust Against Incomplete Fissures

A method for automatic segmentation of pulmonary lobes from computed tomography (CT) scans is presented that is robust against incomplete fissures. The method is based on a multiatlas approach in which existing lobar segmentations are deformed to test scans in which the fissures, the lungs, and the bronchial tree have been automatically segmented. The key element of our method is a cost function that exploits information from fissures, lung borders, and bronchial tree in an effective way, such that less reliable information (lungs, airways) is only used when the most reliable information (fissures) is missing. To cope with the anatomical variation in lobe shape, an atlas selection mechanism is introduced. The method is evaluated on two test sets of 120 scans in total. The results show that the lobe segmentation closely follows the fissures when they are present. In a simulated experiment in which parts of complete fissures are removed, the robustness of the method against different levels of incomplete fissures is shown. When the fissures are incomplete, an observer study shows agreement of the automatically determined lobe borders with a radiologist for 81% of the lobe borders on average.

Towards a close computed tomography monitoring approach for screen detected subsolid pulmonary nodules

Pulmonary subsolid nodules (SSNs) have a high likelihood of malignancy, but are often indolent. A conservative treatment approach may therefore be suitable. The aim of the current study was to evaluate whether close follow-up of SSNs with computed tomography may be a safe approach. The study population consisted of participants of the Dutch-Belgian lung cancer screening trial (Nederlands Leuvens Longkanker Screenings Onderzoek; NELSON). All SSNs detected during the trial were included in this analysis. Retrospectively, all persistent SSNs and SSNs that were resected after first detection were segmented using dedicated software, and maximum diameter, volume and mass were measured. Mass doubling time (MDT) was calculated. In total 7135 volunteers were included in the current analysis. 264 (3.3%) SSNs in 234 participants were detected during the trial. 147 (63%) of these SSNs in 126 participants disappeared at follow-up, leaving 117 persistent or directly resected SSNs in 108 (1.5%) participants available for analysis. The median follow-up time was 95 months (range 20–110 months). 33 (28%) SSNs were resected and 28 of those were (pre-) invasive. None of the non-resected SSNs progressed into a clinically relevant malignancy. Persistent SSNs rarely developed into clinically manifest malignancies unexpectedly. Close follow-up with computed tomography may be a safe option to monitor changes. Persistent subsolid pulmonary nodules may be safely monitored with follow-up computed tomography http://ow.ly/CqWN1

Toward automatic regional analysis of pulmonary function using inspiration and expiration thoracic CT

PURPOSE To analyze pulmonary function using a fully automatic technique which processes pairs of thoracic CT scans acquired at breath-hold inspiration and expiration, respectively. The following research objectives are identified to: (a) describe and systematically analyze the processing pipeline and its results; (b) verify that the quantitative, regional ventilation measurements acquired through CT are meaningful for pulmonary function analysis; (c) identify the most effective of the calculated measurements in predicting pulmonary function; and (d) demonstrate the potential of the system to deliver clinically important information not available through conventional spirometry. METHODS A pipeline of automatic segmentation and registration techniques is presented and demonstrated on a database of 216 subjects well distributed over the various stages of COPD (chronic obstructive pulmonary disorder). Lungs, fissures, airways, lobes, and vessels are automatically segmented in both scans and the expiration scan is registered with the inspiration scan using a fully automatic nonrigid registration algorithm. Segmentations and registrations are examined and scored by expert observers to analyze the accuracy of the automatic methods. Quantitative measures representing ventilation are computed at every image voxel and analyzed to provide information about pulmonary function, both globally and on a regional basis. These CT derived measurements are correlated with results from spirometry tests and used as features in a kNN classifier to assign COPD global initiative for obstructive lung disease (GOLD) stage. RESULTS The steps of anatomical segmentation (of lungs, lobes, and vessels) and registration in the workflow were shown to perform very well on an individual basis. All CT-derived measures were found to have good correlation with spirometry results, with several having correlation coefficients, r, in the range of 0.85-0.90. The best performing kNN classifier succeeded in classifying 67% of subjects into the correct COPD GOLD stage, with a further 29% assigned to a class neighboring the correct one. CONCLUSIONS Pulmonary function information can be obtained from thoracic CT scans using the automatic pipeline described in this work. This preliminary demonstration of the system already highlights a number of points of clinical importance such as the fact that an inspiration scan alone is not optimal for predicting pulmonary function. It also permits measurement of ventilation on a per lobe basis which reveals, for example, that the condition of the lower lobes contributes most to the pulmonary function of the subject. It is expected that this type of regional analysis will be instrumental in advancing the understanding of multiple pulmonary diseases in the future.

Computer-aided Detection of Lung Cancer on Chest Radiographs: Effect on Observer Performance

PURPOSE To assess how computer-aided detection (CAD) affects reader performance in detecting early lung cancer on chest radiographs. MATERIALS AND METHODS In this ethics committee-approved study, 46 individuals with 49 computed tomographically (CT)-detected and histologically proved lung cancers and 65 patients without nodules at CT were retrospectively included. All subjects participated in a lung cancer screening trial. Chest radiographs were obtained within 2 months after screening CT. Four radiology residents and two experienced radiologists were asked to identify and localize potential cancers on the chest radiographs, first without and subsequently with the use of CAD software. A figure of merit was calculated by using free-response receiver operating characteristic analysis. RESULTS Tumor diameter ranged from 5.1 to 50.7 mm (median, 11.8 mm). Fifty-one percent (22 of 49) of lesions were subtle and detected by two or fewer readers. Stand-alone CAD sensitivity was 61%, with an average of 2.4 false-positive annotations per chest radiograph. Average sensitivity was 63% for radiologists at 0.23 false-positive annotations per chest radiograph and 49% for residents at 0.45 false-positive annotations per chest radiograph. Figure of merit did not change significantly for any of the observers after using CAD. CAD marked between five and 16 cancers that were initially missed by the readers. These correctly CAD-depicted lesions were rejected by radiologists in 92% of cases and by residents in 77% of cases. CONCLUSION The sensitivity of CAD in identifying lung cancers depicted with CT screening was similar to that of experienced radiologists. However, CAD did not improve cancer detection because, especially for subtle lesions, observers were unable to sufficiently differentiate true-positive from false-positive annotations.