Kazuto Ashizawa

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.

High resolution CT anatomy of the pulmonary fissures.

Rationale and Objectives: Pulmonary interlobar fissures are important landmarks for proper identification of normal pulmonary anatomy and evaluation of disease. The purpose of this study was to define the radiologic anatomy of the pulmonary fissures using high resolution computed tomography (HRCT) in a large population. Methods: HRCT of the lungs from aortic arch to diaphragm was performed in 622 patients, with a slice thickness of 1 mm and slice interval of 10 mm. Major, minor, and accessory fissures were studied for their orientation and completeness. Results: Both major fissures were mostly facing laterally in their upper parts (100% and 89% right and left, respectively). The left major fissure faced medially (69%) while the right major fissure faced lateral (60%) in their lower parts. The right major fissure was more often incomplete (48% as compared with 43% on the left, P < 0.05). Minor fissures were convex superiorly with the apex in the anterolateral part of the base of the upper lobe, and were incomplete in 63% of cases. Azygos, inferior accessory, superior accessory, and left minor fissures were also seen in 1.2%, 8.6%, 4.6%, and 6.1% of the cases, respectively. Conclusion: The pulmonary fissures are highly variable and the right major fissure differs considerably from the left. The fissures are often incomplete.

Artificial neural networks in chest radiography: application to the differential diagnosis of interstitial lung disease.

RATIONALE AND OBJECTIVES The authors evaluated the usefulness of artificial neural networks (ANNs) in the differential diagnosis of interstitial lung disease. MATERIALS AND METHODS The authors used three-layer, feed-forward ANNs with a back-propagation algorithm. The ANNs were designed to distinguish between 11 interstitial lung diseases on the basis of 10 clinical parameters and 16 radiologic findings extracted by chest radiologists. Thus, the ANNs consisted of 26 input units and 11 output units. One hundred fifty actual clinical cases, 110 cases from previously published articles, and 110 hypothetical cases were used for training and testing the ANNs by using a round-robin (or leave-one-out) technique. ANN performance was evaluated with receiver operating characteristic (ROC) analysis. RESULTS The Az (area under the ROC curve) obtained with actual clinical cases was 0.947, and both the sensitivity and specificity of the ANNs were approximately 90% in terms of indicating the correct diagnosis with the two largest output values among the 11 diseases. CONCLUSION ANNs using clinical parameters and radiologic findings may be useful for making the differential diagnosis of interstitial lung disease on chest radiographs.

Artificial neural networks (ANNs) for differential diagnosis of interstitial lung disease : results of a simulation test with actual clinical cases1

Rationale and Objectives. To evaluate the performance of an artificial neural network (ANN) scheme with use of consecutive clinical cases and its effect on radiologists with an observer test. Materials and Methods. Artificial neural networks were designed to distinguish among 11 interstitial lung diseases on the basis of 26 inputs (16 radiologic findings, 10 clinical parameters). Chest radiographs of 96 consecutive cases with interstitial lung disease were used. Five radiologists independently rated their radiologic findings on the 96 chest radiographs. Based on their ratings of radiologic findings and clinical parameters obtained from the hospital information system, the output values indicating the likelihood of each of the 11 interstitial lung diseases were determined. Subsequently, 30 cases were selected from these 96 cases for an observer test. Five radiologists marked their confidence levels for diagnosis of 11 possible diseases in each case without and with ANN output. The performance of ANNs and radiologists was evaluated by receiver operating characteristic analysis based on their outputs and on confidence levels, respectively. Results. The average AZ value (area under the receiver operating characteristic curve) indicating ANN performance for the 96 consecutive cases was 0.85 ± 0.03. The average AZ values indicating radiologists’ performance without and with ANN outputs were 0.81 ± 0.11 and 0.87 ± 0.06, respectively. The diagnostic accuracy was improved significantly when radiologists read chest radiographs with ANN outputs (P <.05). Conclusion. Artificial neural networks for differential diagnosis of interstitial lung disease may be useful in clinical situations, and radiologists may be able to utilize the ANN output to their advantage in the differential diagnosis of interstitial lung disease on chest radiographs.

Measurement of Focal Ground-glass Opacity Diameters on CT Images: Interobserver Agreement in Regard to Identifying Increases in the Size of Ground-Glass Opacities

PURPOSE To evaluate interobserver agreement in regard to measurements of focal ground-glass opacities (GGO) diameters on computed tomography (CT) images to identify increases in the size of GGOs. MATERIALS AND METHODS Approval by the institutional review board and informed consent by the patients were obtained. Ten GGOs (mean size, 10.4 mm; range, 6.5-15 mm), one each in 10 patients (mean age, 65.9 years; range, 58-78 years), were used to make the diameter measurements. Eleven radiologists independently measured the diameters of the GGOs on a total of 40 thin-section CT images (the first [n = 10], the second [n = 10], and the third [n = 10] follow-up CT examinations and remeasurement of the first [n = 10] follow-up CT examinations) without comparing time-lapse CT images. Interobserver agreement was assessed by means of Bland-Altman plots. RESULTS The smallest range of the 95% limits of interobserver agreement between the members of the 55 pairs of the 11 radiologists in regard to maximal diameter was -1.14 to 1.72 mm, and the largest range was -7.7 to 1.7 mm. The mean value of the lower limit of the 95% limits of agreement was -3.1 ± 1.4 mm, and the mean value of their upper limit was 2.5 ± 1.1 mm. CONCLUSION When measurements are made by any two radiologists, an increase in the length of the maximal diameter of more than 1.72 mm would be necessary in order to be able to state that the maximal diameter of a particular GGO had actually increased.

Quantification of lung perfusion blood volume (lung PBV) by dual-energy CT in patients with and without pulmonary embolism: preliminary results.

OBJECTIVES Recently, software has been used for quantification of lung PBV, which can be evaluated objectively; however, this technique is yet to be validated. The purpose was to investigate the clinical feasibility of the quantification of lung perfusion blood volume (lung PBV) by dual-energy CT in patients with pulmonary embolism (PE). MATERIALS AND METHODS One hundred thirteen patients with clinical suspicion of PE underwent dual-energy CT angiography with a standard injection protocol. Patients were divided into each two groups with and without PE based on the presence of endoluminal clots on transverse diagnostic scans. We evaluated the quantification of lung PBV using a workstation. Associations between lung PVB and the numbers of pulmonary segments with PE were also evaluated. RESULTS Thirty three of 113 (29%) patients were found to have endoluminal clots in the right and/or left lungs. The remaining 80 patients did not have endoluminal clots. In 33 patients, the mean number of segments with endoluminal clots was 5.2±3.3. For patient (whole lung)-based analysis, in patients with and without PE, mean lung PBVs were 20.8±2.3 and 28.7±6.8 Hounsfield Unit (HU), respectively, with a significant difference between the two groups (p<0.0001). In patients with PE, there was a significant correlation between lung PBV and the numbers of pulmonary segments with PE (R=0.57, p=0.0005). CONCLUSION The findings of this preliminary study suggest that quantification of lung PBV may reflect the pulmonary artery perfusion, which is useful to evaluate pulmonary blood flow in patients with PE.

Use of an Artificial Neural Network to Determine the Diagnostic Value of Specific Clinical and Radiologic Parameters in the Diagnosis of Interstitial Lung Disease on Chest Radiographs

RATIONALE AND OBJECTIVES The authors investigated the diagnostic value of each of multiple clinical parameters and radiologic findings in differentiating between various interstitial lung diseases by using an artificial neural network (ANN). MATERIALS AND METHODS The ANN was designed to differentiate between 11 interstitial lung diseases. The authors employed 10 clinical parameters and 16 radiologic findings that were divided into three groups (location, general appearance, specific findings). The performance of the ANN was evaluated with receiver operating characteristic analysis with amodified round-robin (leave-one-out) method and 370 cases (150 actual cases, 110 published cases, and 110 hypothetical cases). The Az values of ANNs were evaluated with various combinations of 10 clinical parameters and 16 radiologic findings. RESULTS The Az value obtained with the complete set of clinical parameters and radiologic findings was 0.947. The Az value obtained with the 10 clinical parameters alone was 0.900, which was greater than 0.843 obtained with the 16 radiologic findings alone. There were statistically significant differences among Az values for some diseases when certain clinical parameters were removed from the input. Omission of specific findings among the three groups of radiologic findings decreased the Az value significantly. CONCLUSION These results appear to confirm that clinical parameters can be equally as or more important than radiologic findings in the diagnosis of interstitial lung diseases. Among radiologic findings, certain specific findings can be more important than the location or general appearance of abnormal findings.

Anomalous systemic arterial supply to normal basal segments of left lower lobe: characteristic imaging findings.

Purpose The purpose of this study was to describe the characteristic findings in each imaging modality of anomalous systemic arterial supply to normal basal segments of the left lower lobe of the lung. Method Five patients with anomalous systemic arterial supply to normal basal segments of the left lower lobe were retrospectively reviewed. Chest radiography, contrast-enhanced CT scan, angiography, and other imaging modalities were analyzed. Results The imaging findings of this anomaly were characterized by three issues as follows: an anomalous systemic artery arising from the thoracic aorta, absence of pulmonary arterial supply, and normal bronchial system as well as normal pulmonary parenchyma in the affected segments. The anomalous systemic artery was demonstrated on chest radiography, contrast-enhanced CT scan, MRI, and thoracic aortography. Absence of pulmonary arterial supply was revealed by pulmonary angiography, radiolabeled perfusion scan, and CT scan. Normal bronchial system was confirmed by radiolabeled ventilation scan, bronchography, and CT scan. Conclusion CT is useful in making the correct diagnosis of this anomaly because it is the only diagnostic method that can demonstrate the anomalous systemic artery, absence of pulmonary arterial supply, and normal bronchial system in the affected segments.

Natural History of Pulmonary Subsolid Nodules: A Prospective Multicenter Study

Introduction: The purpose of this study was to evaluate the natural course of the progression of pulmonary subsolid nodules (SSNs). Materials and Methods: Eight facilities participated in this study. A total of 795 patients with 1229 SSNs were assessed for the frequency of invasive adenocarcinomas. SSNs were classified into three categories: pure ground‐glass nodules (PGGNs), heterogeneous GGNs (HGGNs) (solid component detected only in lung windows), and part‐solid nodules. Results: The mean prospective follow‐up period was 4.3 ± 2.5 years. SSNs were classified at baseline as follows: 1046 PGGNs, 81 HGGNs, and 102 part‐solid nodules. Among the 1046 PGGNs, 13 (1.2%) developed into HGGNs and 56 (5.4%) developed into part‐solid nodules. Among the 81 HGGNs, 16 (19.8%) developed into part‐solid nodules. Thus, the SSNs at the final follow‐up were classified as follows: 977 PGGNs, 78 HGGNs, and 174 part‐solid nodules. Of the 977 PGGNs, 35 were resected (nine minimally invasive adenocarcinomas [MIAs], 21 adenocarcinomas in situ [AIS], and five atypical adenomatous hyperplasias). Of the 78 HGGNs, seven were resected (five MIAs and two AIS). Of the 174 part‐solid nodules, 49 were resected (12 invasive adenocarcinomas, 26 MIAs, 10 AIS, and one adenomatous hyperplasia). For the PGGNs, the mean period until their development into part‐solid nodules was 3.8 ± 2.0 years, whereas the mean period for the HGGNs was 2.1 ± 2.3 years (p = 0.0004). Conclusion: This study revealed the frequencies and periods of development from PGGNs and HGGNs into part‐solid nodules. Invasive adenocarcinomas were diagnosed only among the part‐solid nodules, corresponding to 1% of all 1229 SSNs.

Application of Artificial Neural Networks for Quantitative Analysis of Image Data in Chest Radiographs for Detection of Interstitial Lung Disease

The authors have developed an automated computeraided diagnostic (CAD) scheme by using artificial neural networks (ANNs) on quantitative analysis of image data. Three separate ANNs were applied for detection of interstitial disease on digitized chest images. The first ANN was trained with horizontal profiles in regions of interest (ROIs) selected from normal and abnormal chest radiographs for distinguishing between normal and abnormal patterns. For training and testing of the second ANN, the vertical output patterns obtained from the 1st ANN were used for each ROI. The output value of the second ANN was used to distinguish between normal and abnormal ROIs with interstitial infiltrates. If the ratio of the number of abnormal ROIs to the total number of all ROIs in a chest image was greater than a specified threshold level, the image was classified as abnormal. In addition, the third ANN was applied to distinguish between normal and abnormal chest images. The combination of the rule-based method and the third ANN also was applied to the classification between normal and abnormal chest images. The performance of the ANNs was evaluated by means of receiver operating characteristic (ROC) analysis. The average Az value (area under the ROC curve) for distinguishing between normal and abnormal cases was 0.976±0.012 for 100 chest radiographs that were not used in training of ANNs. The results indicate that the ANN trained with image data can learn some statistical properties associated with interstitial infiltrates in chest radiographs.