Computerized analysis of abnormal asymmetry in digital chest radiographs: evaluation of potential utility.
The purpose of this study was to develop and test a computerized method for the fully automated analysis of abnormal asymmetry in digital posteroanterior (PA) chest radiographs. An automated lung segmentation method was used to identify the aerated lung regions in 600 chest radiographs. Minimal a priori lung morphology information was required for this gray-level thresholding-based segmentation. Consequently, segmentation was applicable to grossly abnormal cases. The relative areas of segmented right and left lung regions in each image were compared with the corresponding area distributions of normal images to determine the presence of abnormal asymmetry. Computerized diagnoses were compared with image ratings assigned by a radiologist. The ability of the automated method to distinguish normal from asymmetrically abnormal cases was evaluated by using receiver operating characteristic (ROC) analysis, which yielded an area under the ROC curve of 0.84. This automated method demonstrated promising performance in its ability to detect abnormal asymmetry in PA chest images. We believe this method could play a role in a picture archiving and communications (PACS) environment to immediately identify abnormal cases and to function as one component of a multifaceted computer-aided diagnostic scheme. (+info)
A reliability study for evaluating information extraction from radiology reports.
GOAL: To assess the reliability of a reference standard for an information extraction task. SETTING: Twenty-four physician raters from two sites and two specialties judged whether clinical conditions were present based on reading chest radiograph reports. METHODS: Variance components, generalizability (reliability) coefficients, and the number of expert raters needed to generate a reliable reference standard were estimated. RESULTS: Per-rater reliability averaged across conditions was 0.80 (95% CI, 0.79-0.81). Reliability for the nine individual conditions varied from 0.67 to 0.97, with central line presence and pneumothorax the most reliable, and pleural effusion (excluding CHF) and pneumonia the least reliable. One to two raters were needed to achieve a reliability of 0.70, and six raters, on average, were required to achieve a reliability of 0.95. This was far more reliable than a previously published per-rater reliability of 0.19 for a more complex task. Differences between sites were attributable to changes to the condition definitions. CONCLUSION: In these evaluations, physician raters were able to judge very reliably the presence of clinical conditions based on text reports. Once the reliability of a specific rater is confirmed, it would be possible for that rater to create a reference standard reliable enough to assess aggregate measures on a system. Six raters would be needed to create a reference standard sufficient to assess a system on a case-by-case basis. These results should help evaluators design future information extraction studies for natural language processors and other knowledge-based systems. (+info)
Comparison of low attenuation areas on computed tomographic scans between inner and outer segments of the lung in patients with chronic obstructive pulmonary disease: incidence and contribution to lung function.
BACKGROUND: The low attenuation areas on computed tomographic (CT) scans have been reported to represent emphysematous changes of the lung. However, the regional distribution of emphysema between the inner and outer segments of the lung has not been adequately studied. In this study the regional distribution of low attenuation areas has been compared by quantitative CT analysis and the contribution of the regional distribution to pulmonary function tests evaluated in patients with chronic obstructive pulmonary disease (COPD). METHODS: Chest CT images and the results of pulmonary function tests were obtained from 73 patients with COPD. The lung images were divided into inner and outer segments in the upper (cranial), middle, and lower (caudal) sections. The percentage ratio of low attenuation area to corresponding lung area (LAA%) was then calculated. The LAA% of each segment was also compared with the results of pulmonary function tests. RESULTS: The mean (SD) LAA% of the inner segment was 39.1 (18.5) compared with 28.1 (13.2) for the outer segment (p<0.0001). Linear and multiple regression analyses revealed that airflow limitation is closely correlated with the inner segment LAA% of the lower lung. In contrast, the carbon monoxide transfer factor is closely correlated with the inner segment LAA% of the upper lung. CONCLUSION: Low attenuation areas on CT scans are more often found in the inner segment of the lung than in the outer segment, and the contribution of the inner segment to pulmonary function tests may be greater than the outer segment. (+info)
Contrast-enhanced spiral CT of the head and neck: comparison of contrast material injection rates.
BACKGROUND AND PURPOSE: Contrast-enhanced spiral CT studies of the head and neck are performed frequently using contrast material volumes of approximately 30 g iodine and a scan delay of 30-45 seconds. Because little is known about the effects of contrast material injection rates on tissue enhancement, this was prospectively investigated in our study. METHODS: Ninety-seven patients underwent spiral CT of the head and neck. Each patient was assigned randomly to one of four groups who received 100 mL of nonionic contrast material (300 mg I/mL) at different monophasic injection flow rates with 1.5, 2, 3, and 4 mL/s. Scanning started after a constant delay of 35 seconds. The attenuation of the carotid artery, jugular vein, and sternocleidomastoid muscle was measured over time and the attenuation of the submandibular and thyroid gland was evaluated. Vascular attenuation of at least 150 HU was considered to be sufficient. RESULTS: The mean scan time was 33+/-5 seconds. The study, using an injection rate of 2 mL/s, showed the longest time of sufficient overall (arterial and venous) vessel attenuation (27+/-4 seconds, P< or =.008). The injection flow rate did not influence significantly muscular attenuation (mean enhancement during scan time: 9+/-7 HU). The 1.5 mL/s protocol showed the lowest attenuation values of the submandibular gland (81+/-12 HU) and the highest attenuation values of the thyroid gland (164+/-22 HU), but the attenuation of the thyroid gland was not statistically different from that revealed by the 2 mL/s protocol. CONCLUSION: Using 100 mL of intravenous contrast material with 300 mg I/mL for spiral CT studies of the entire head and neck, the optimal injection flow is 2 mL/s, whereas lower flow rates resulted in insufficient venous enhancement. (+info)
Comparing expert systems for identifying chest x-ray reports that support pneumonia.
We compare the performance of four computerized methods in identifying chest x-ray reports that support acute bacterial pneumonia. Two of the computerized techniques are constructed from expert knowledge, and two learn rules and structure from data. The two machine learning systems perform as well as the expert constructed systems. All of the computerized techniques perform better than a baseline keyword search and a lay person, and perform as well as a physician. We conclude that machine learning can be used to identify chest x-ray reports that support pneumonia. (+info)
Fracture interpretation using electronic presentation: a comparison.
The purposes of this study were to determine whether (1) fractures are interpreted differently after digitization and electronic presentation; (2) there are differences in accuracy between screen radiographs and electronic presentation; (3) differences in interpretation are a function of monitor resolution; and (4) differences in interpretation between radiographs and electronic images relate to radiological subspecialty. Forty cases with fractures of varying degrees of subtlety and 35 cases without fractures were interpreted. Radiographs were digitized with 2 different systems and displayed on 3 monitors of different spatial resolution. Four radiologists, with varying experience, were asked to decide whether fractures were present, absent, or they were uncertain. Accuracy of interpretation increased with improved electronic image presentation and monitor resolution. The sensitivity, specificity, and accuracy of fracture detection on System A were 63%, 98%, and 78%, respectively. The results were 72%, 98%, and 84% with System B. System C results were 81%, 97%, and 88% with Lumiscan 75, and 82%, 96%, and 88% with Lumiscan 150. Sensitivity, specificity, and accuracy results of the original radiograph interpretation were 89%, 95%, and 92%. Results were significantly different for System A. No significant differences were found for the other systems compared with film radiographs. System A did not have adequate monitors for interpretation of subtle fractures. Systems B and C were capable of displaying even subtle fractures. Our initial results indicate that interpretation with high-quality 1K x 1K monitors is substantially similar to radiograph interpretation. (+info)
American College of Cardiology/ European Society of Cardiology international study of angiographic data compression phase I. The effects of lossy data compression on recognition of diagnostic features in digital coronary angiography.
OBJECTIVES: This study intended to determine the effect of varying degrees of lossy Joint Photographic Experts Group (JPEG) compression on detection of coronary angiographic features. Background Compression of digital coronary angiograms facilitates playback of images and decreases cost. There are little data on the effect of compression on the accuracy of coronary angiography. METHODS: At six centers, 71 angiographers each reviewed a set of 100 angiographic sequences. The 100 sequences were divided into four, 25-sequence subsets. Each subset of 25 was displayed either as original images or at one of three compression ratios (CRs) (6:1, 10:1 or 16:1). The effect of lossy compression on the sensitivity and specificity for detection of diagnostic features was determined. The effect of compression on subjective measures of image quality graded by the angiographers was also examined. RESULTS: Lossy compression at a ratio of 16:1 decreased the sensitivity for the detection of diagnostic features (76% vs. 80%P=0.004). The largest effect was in the detection of calcification (52% vs. 63% at 16:1 compression vs. original images, P<0.001). Subjective indicators of image quality indicated a reduction in confidence in interpretation at CRs of 10:1 and 16:1. CONCLUSIONS: With increased ratios of lossy compression, a degradation of digital coronary angiograms occurs that results in decreased diagnostic accuracy. The sensitivity for detection of common diagnostic features was decreased, and subjective assessment of image quality was impaired. Caution is warranted in the interpretation of coronary angiograms that have been subjected to lossy JPEG compression beyond a ratio of 6:1. (+info)
American College of Cardiology/ European Society of Cardiology international study of angiographic data compression phase II. The effects of varying JPEG data compression levels on the quantitative assessment of the degree of stenosis in digital coronary angiography.
OBJECTIVES: This report describes whether lossy Joint Photographic Experts Group (JPEG) image compression/decompression has an effect on the quantitative assessment of vessel sizes by state-of-the-art quantitative coronary arteriography (QCA). BACKGROUND: The Digital Imaging and Communications in Medicine (DICOM) digital exchange standard for angiocardiography prescribes that images must be stored loss free, thereby limiting JPEG compression to a maximum ratio of 2:1. For practical purposes it would be desirable to increase the compression ratio (CR), which would lead to lossy image compression. METHODS: A series of 48 obstructed coronary segments were compressed/decompressed at CR 1:1 (uncompressed), 6:1, 10:1 and 16:1 and analyzed blindly and in random order using the QCA-CMS analytical software. Similar catheter and vessel start- and end-points were used within each image quartet, respectively. All measurements were repeated after several weeks using newly selected start- and end-points. Three different sub-analyses were carried out: the intra-observer, fixed inter-compression and variable inter-compression analyses, with increasing potential error sources, respectively. RESULTS: The intra-observer analysis showed significant systematic and random errors in the calibration factor at JPEG CR 10:1. The fixed inter-compression analysis demonstrated systematic errors in the calibration factor and recalculated vessel parameter results at CR 16:1 and for the random errors at CR 10:1 and 16:1. The variable inter-compression analysis presented systematic and random errors in the calibration factor and recalculated parameter results at CR 10:1 and 16:1. Any negative effect at CR 6:1 was found only for the calibration factor of the variable inter-compression analysis, which did not show up in the final vessel measurements. CONCLUSIONS: Compression ratios of 10:1 and 16:1 affected the QCA results negatively and therefore should not be used in clinical research studies. (+info)