Roger Engelmann

Computer-aided diagnostic scheme for the detection of lung nodules on chest radiographs: localized search method based on anatomical classification.

We developed an advanced computer-aided diagnostic (CAD) scheme for the detection of various types of lung nodules on chest radiographs intended for implementation in clinical situations. We used 924 digitized chest images (992 noncalcified nodules) which had a 500 x 500 matrix size with a 1024 gray scale. The images were divided randomly into two sets which were used for training and testing of the computerized scheme. In this scheme, the lung field was first segmented by use of a ribcage detection technique, and then a large search area (448 x 448 matrix size) within the chest image was automatically determined by taking into account the locations of a midline and a top edge of the segmented ribcage. In order to detect lung nodule candidates based on a localized search method, we divided the entire search area into 7 x 7 regions of interest (ROIs: 64 x 64 matrix size). In the next step, each ROI was classified anatomically into apical, peripheral, hilar, and diaphragm/heart regions by use of its image features. Identification of lung nodule candidates and extraction of image features were applied for each localized region (128 x 128 matrix size), each having its central part (64 x 64 matrix size) located at a position corresponding to a ROI that was classified anatomically in the previous step. Initial candidates were identified by use of the nodule-enhanced image obtained with the average radial-gradient filtering technique, in which the filter size was varied adaptively depending on the location and the anatomical classification of the ROI. We extracted 57 image features from the original and nodule-enhanced images based on geometric, gray-level, background structure, and edge-gradient features. In addition, 14 image features were obtained from the corresponding locations in the contralateral subtraction image. A total of 71 image features were employed for three sequential artificial neural networks (ANNs) in order to reduce the number of false-positive candidates. All parameters for ANNs, i.e., the number of iterations, slope of sigmoid functions, learning rate, and threshold values for removing the false positives, were determined automatically by use of a bootstrap technique with training cases. We employed four different combinations of training and test image data sets which was selected randomly from the 924 cases. By use of our localized search method based on anatomical classification, the average sensitivity was increased to 92.5% with 59.3 false positives per image at the level of initial detection for four different sets of test cases, whereas our previous technique achieved an 82.8% of sensitivity with 56.8 false positives per image. The computer performance in the final step obtained from four different data sets indicated that the average sensitivity in detecting lung nodules was 70.1% with 5.0 false positives per image for testing cases and 70.4% sensitivity with 4.2 false positives per image for training cases. The advanced CAD scheme involving the localized search method with anatomical classification provided improved detection of pulmonary nodules on chest radiographs for 924 lung nodule cases.

Application of temporal subtraction for detection of interval changes on chest radiographs: Improvement of subtraction images using automated initial image matching

The authors developed a temporal subtraction scheme based on a nonlinear geometric warping technique to assist radiologists in the detection of interval changes in chest radiographs obtained on different occasions. The performance of the current temporal subtraction scheme is reasonably good; however, severe misregistration can occur in some cases. The authors evaluated the quality of 100 chest temporal subtraction images selected from their clinical image database. Severe misregistration was mainly attributable to initial incorrect global matching. Therefore, they attempted to improve the quality of the subtraction images by applying a new initial image matching technique to determine the global shift value between the current and the previous chest images. A cross-correlation method was employed for the initial image matching by use of blurred low-resolution chest images. Nineteen cases (40.4%) among 47 poor registered subtraction images were improved. These results show that the new initial image matching technique is very effective for improving the quality of chest temporal subtraction images, which can greatly enhance subtle changes in chest radiographs.

Dual energy subtraction and temporal subtraction chest radiography.

Digital radiography and display systems have revolutionized radiologic practice in recent years and have enabled clinical application of advanced image processing techniques. These include dual energy subtraction and temporal subtraction, both of which can improve diagnostic accuracy for abnormal findings in chest radiographs, especially for subtle lesions such as early lung cancer or focal pneumonia. Dual energy radiography exploits the differential attenuation of low-energy x-ray photons by calcium to produce separate images on the bones and soft tissues, which provides improved detection and characterization of both calcified and noncalcified lung lesions. Dual energy subtraction radiography is currently available from 2 of the major vendors and is in clinical use at many institutions in the United States. Temporal subtraction is a complementary technique that enhances interval change, by using a previous radiograph as a subtraction mask, so that unchanged normal anatomy is suppressed, whereas new abnormalities are enhanced. Though it is not yet a product in the United States, temporal subtraction is available for clinical use in Japan. Temporal subtraction can be combined with energy subtraction to reduce misregistration artifacts, and also has potential to improve computer-aided detection of nodules and other types of lung disease.

Usefulness of Texture Analysis for Computerized Classification of Breast Lesions on Mammograms

This work presents the usefulness of texture features in the classification of breast lesions in 5518 images of regions of interest, which were obtained from the Digital Database for Screening Mammography that included microcalcifications, masses, and normal cases. Sixteen texture features were used, i.e., 13 were based on the spatial gray-level dependence matrix and 3 on the wavelet transform. The nonparametric K-NN classifier was used in the classification stage. The results obtained from receiver operating characteristic analysis indicated that the texture features can be used for separating normal regions and lesions with masses and microcalcifications, yielding the area under the curve (AUC) values of 0.957 and 0.859, respectively. However, the texture features were not very effective for distinguishing between malignant and benign lesions because the AUC was 0.617 for masses and 0.607 for microcalcifications. The study showed that the texture features can be used for the detection of suspicious regions in mammograms.

Effect of High Sensitivity in a Computerized Scheme for Detecting Extremely Subtle Solitary Pulmonary Nodules in Chest Radiographs: Observer Performance Study

RATIONALES AND OBJECTIVES This study investigated the effect of a high sensitivity in computer-aided diagnosis (CAD) for detecting lung nodules in chest radiographs when extremely subtle cases were presented to radiologists. MATERIAL AND METHODS The chest radiographs used in this study consisted of 36 normal images and 54 abnormals containing solitary lung nodules, of which 25 were extremely subtle and 29 were very subtle. Receiver operating characteristic analysis for detecting lung nodules was performed without and with CAD. The levels of CAD output were simulated with a hypothetical ideal performance of 100% sensitivity, but with three or four false positives per image. Six radiologists participated in an observer study in which cases were interpreted first without and then with the use of CAD. RESULTS The average A(z) values for radiologists without and with CAD were 0.682 and 0.808, respectively. The performance of radiologists was improved significantly when high sensitivity was used (P = .0003). However, the radiologists were not able to recognize some extremely subtle nodules (5 of 54 nodules by all radiologists), even with the correct CAD output; these nodules were then considered as non-actionable. None of 306 computer-false positives was incorrectly regarded as a nodule by all radiologists, but 63 false positives were incorrectly identified by one or more radiologists. CONCLUSION The accuracy of radiologists in the detection of some extremely subtle solitary pulmonary nodules can be improved significantly when the sensitivity of a CAD scheme can be made to be at an extremely high level. However, all of the six radiologists failed to identify some nodules (about 10%), even with the correct output of the CAD.

Temporal subtraction of dual-energy chest radiographs.

Temporal subtraction and dual-energy imaging are two enhanced radiography techniques that are receiving increased attention in chest radiography. Temporal subtraction is an image processing technique that facilitates the visualization of pathologic change across serial chest radiographic images acquired from the same patient; dual-energy imaging exploits the differential relative attenuation of x-ray photons exhibited by soft-tissue and bony structures at different x-ray energies to generate a pair of images that accentuate those structures. Although temporal subtraction images provide a powerful mechanism for enhancing visualization of subtle change, misregistration artifacts in these images can mimic or obscure abnormalities. The purpose of this study was to evaluate whether dual-energy imaging could improve the quality of temporal subtraction images. Temporal subtraction images were generated from 100 pairs of temporally sequential standard radiographic chest images and from the corresponding 100 pairs of dual-energy, soft-tissue radiographic images. The registration accuracy demonstrated in the resulting temporal subtraction images was evaluated subjectively by two radiologists. The registration accuracy of the soft-tissue-based temporal subtraction images was rated superior to that of the conventional temporal subtraction images. Registration accuracy also was evaluated objectively through an automated method, which achieved an area-under-the-ROC-curve value of 0.92 in the distinction between temporal subtraction images that demonstrated clinically acceptable and clinically unacceptable registration accuracy. By combining dual-energy soft-tissue images with temporal subtraction, misregistration artifacts can be reduced and superior image quality can be obtained.

Improved detection of focal pneumonia by chest radiography with bone suppression imaging

AbstractObjectiveTo evaluate radiologists’ ability to detect focal pneumonia by use of standard chest radiographs alone compared with standard plus bone-suppressed chest radiographs.MethodsStandard chest radiographs in 36 patients with 46 focal airspace opacities due to pneumonia (10 patients had bilateral opacities) and 20 patients without focal opacities were included in an observer study. A bone suppression image processing system was applied to the 56 radiographs to create corresponding bone suppression images. In the observer study, eight observers, including six attending radiologists and two radiology residents, indicated their confidence level regarding the presence of a focal opacity compatible with pneumonia for each lung, first by use of standard images, then with the addition of bone suppression images. Receiver operating characteristic (ROC) analysis was used to evaluate the observers’ performance.ResultsThe mean value of the area under the ROC curve (AUC) for eight observers was significantly improved from 0.844 with use of standard images alone to 0.880 with standard plus bone suppression images (P < 0.001) based on 46 positive lungs and 66 negative lungs.ConclusionUse of bone suppression images improved radiologists’ performance for detection of focal pneumonia on chest radiographs.Key Points• Bone suppression image processing can be applied to conventional digital radiography systems. • Bone suppression imaging (BSI) produces images that appear similar to dual-energy soft tissue images. • BSI improves the conspicuity of focal lung disease by minimizing bone opacity. • BSI can improve the accuracy of radiologists in detecting focal pneumonia.

Clinical utility of temporal subtraction images in successive whole-body bone scans: Evaluation in a prospective clinical study

In order to aid radiologists’ routine work for interpreting bone scan images, we developed a computerized method for temporal subtraction (TS) images which can highlight interval changes between successive whole-body bone scans, and we performed a prospective clinical study for evaluating the clinical utility of the TS images. We developed a TS image server which includes an automated image-retrieval system, an automated image-conversion system, an automated TS image-producing system, a computer interface for displaying and evaluating TS images with five subjective scales, and an automated data-archiving system. In this study, the radiologist could revise his/her report after reviewing the TS images if the findings on the TS image were confirmed retrospectively on our clinical picture archiving and communication system. We had 256 consenting patients of whom 143 had two or more whole-body bone scans available for TS images. In total, we obtained TS images successfully in 292 (96.1%) pairs and failed to produce TS images in 12 pairs. Among the 292 TS studies used for diagnosis, TS images were considered as “extremely beneficial” or “somewhat beneficial” in 247 (84.6%) pairs, as “no utility” in 44 pairs, and as “somewhat detrimental” in only one pair. There was no TS image for any pairs that was considered “extremely detrimental.” In addition, the radiologists changed their initial reported impression in 18 pairs (6.2%). The benefit to the radiologist of using TS images in the routine interpretation of successive whole-body bone scans was significant, with negligible detrimental effects.

Effect of the number of cases in image database on the performance of computer-aided diagnosis (CAD) for the detection of pulmonary nodules in chest radiographs

We investigated the effect of the number of cases included in an image database on development of a computer-aided diagnosis (CAD) scheme for the detection of lung nodules, in terms of the performance of the CAD scheme. A total number of 1000 chest radiographs with nodules was used in this study. All images were divided randomly into subsets consisting of the same number of cases from different sources. The subsets we used in this study were 10 sets of 100 cases, 5 sets of 200 cases, and 2 sets of 500 cases. The entire database and all of the subsets were tested by use of the same CAD scheme, but with different parameter settings for consistency tests. When the sensitivities of the CAD scheme for each subset were kept at a level of 70.0 %, the numbers of false positives per image were 0.1 for 100 cases, 0.6 for 200 cases, 2.9 for 500 cases, and 6.2 for 1000 cases. Therefore, the performance of the CAD scheme in detecting lung nodules was strongly affected by the number of cases used. We conclude that a large-scale image database is needed for reliable evaluation of the performance of CAD.

LUNGx Challenge for computerized lung nodule classification

Abstract. The purpose of this work is to describe the LUNGx Challenge for the computerized classification of lung nodules on diagnostic computed tomography (CT) scans as benign or malignant and report the performance of participants’ computerized methods along with that of six radiologists who participated in an observer study performing the same Challenge task on the same dataset. The Challenge provided sets of calibration and testing scans, established a performance assessment process, and created an infrastructure for case dissemination and result submission. Ten groups applied their own methods to 73 lung nodules (37 benign and 36 malignant) that were selected to achieve approximate size matching between the two cohorts. Area under the receiver operating characteristic curve (AUC) values for these methods ranged from 0.50 to 0.68; only three methods performed statistically better than random guessing. The radiologists’ AUC values ranged from 0.70 to 0.85; three radiologists performed statistically better than the best-performing computer method. The LUNGx Challenge compared the performance of computerized methods in the task of differentiating benign from malignant lung nodules on CT scans, placed in the context of the performance of radiologists on the same task. The continued public availability of the Challenge cases will provide a valuable resource for the medical imaging research community.