| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Department of Diagnostic Imaging, University of Maryland Medical System, 22 S.
Greene St., Baltimore, MD 21201.
2
Department of Diagnostic Imaging, Baltimore Veterans Affairs Medical Center,
10 N. Greene St., Baltimore, MD 21201.
3
Present address: American Radiology Associates, 1838 Green Tree Rd.,
Baltimore, MD 21208
4
Present address: Department of Diagnostic Radiology, South Manchester
University Hospitals NHS Trust, Wythenshawe Hospital, S. Moor Rd., Manchester
M23 9LT, United Kingdom.
Received April 14, 1999;
accepted after revision July 7, 1999.
Address correspondence to S. M. Pomerantz.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Two hundred three randomly selected abdominal and chest CT scans were interpreted by three radiologists using a four-monitor soft-copy workstation (images per screen, nine; resolution, 2K). After the initial interpretations, all scans were reevaluated by the same radiologists using additional liver and bone window and level settings. Differences in conspicuity and characterization of abnormalties were graded on a three-point scale.
RESULTS. Conspicuity and characterization of abnormalities were improved in 67% of abnormal findings (81/121; p = 0.01). Improvement (a finding that substantially affected the final diagnosis) was present in 18% of abnormal findings (22/121; p = 0.04). On average, the evaluation of images at multiple window and level settings required an additional 40 sec per case.
CONCLUSION. The use of multiple window and level settings during soft-copy interpretation resulted in improved lesion detectability and characterization with greater diagnostic efficacy. Using soft-copy workstations, radiologists can evaluate images using multiple settings without compromising efficiency.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We reviewed the radiology literature and found few reports examining the use and cost-effectiveness of multiple organ-specific window and level settings on CT. To our knowledge, the subject has been partially explored in three studies; however, each study was performed in a film-based environment [6, 7, 8]. Nevertheless, conventional wisdom among radiologists suggests that the use of additional settings during CT interpretation increases diagnostic accuracy [9, 10, 11, 12, 13]. For example, soft-tissue, liver, and bone settings are designed to show specific structures to the best advantage. To view multiple window and level settings in a film-based environment, each image set must be individually printed. The printing of additional sequences increases costs and decreases efficiency for technologists, film library personnel, and radiologists. In practice, especially given the current medical-economic pressures, these factors limit the number of window and level settings that can be used for any CT scan. For example, at our institution, typical chest CT is filmed with soft-tissue and lung windows unless another setting is specifically indicated. Given the ease of manipulating window and level settings during soft-copy interpretation, the routine use of additional settings (such as liver windows for the upper abdomen) might increase diagnostic accuracy without requiring additional time. For abdominal CT, the use of liver and bone settings may increase diagnostic accuracy with minimum additional effort. In this study, we investigated the value and efficiency of using multiple window and level settings for chest and abdominal CT during soft-copy interpretation.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Initially, each chest CT scan was interpreted with lung (window width, 2000 H; window level, -700 H) and soft-tissue windows (window width, 400 H; window level, 40 H). Each abdominal CT was interpreted with soft-tissue windows (window width, 400 H; window level, 40 H). After interpretation, the radiologists were asked to note all findings related to the liver and skeletal structures. Immediately after the initial interpretation, liver (window width, 350 H; window level, 50 H) and bone (window width, 1600 H; window level, 400 H) settings were applied sequentially. The radiologists were then asked to note any change in interpretation. Small degenerative bone changes were considered normal for the purposes of this investigation.
Changes in diagnosis after the second interpretation were rated on a three-point scale. Type 1 images revealed no new findings and did not change the original diagnosis. Type 2 images revealed new findings of doubtful importance or illustrated increased conspicuity or characterization of previously identified findings of doubtful importance. Type 3 images revealed new findings that substantially affected the final diagnosis or illustrated increased conspicuity or improved characterization of findings that substantially affected the final diagnosis.
To guide the judgment of clinical importance, radiologists were given several examples. The identification of a liver or bone lesion overlooked on the original settings would be a substantial finding of clinical importance (type 3). The identification of additional lesions when numerous lesions were previously identified would not be considered clinically important (type 2). Bone abnormalities initially presumed to represent degenerative change on the original lung or soft-tissue settings that later were considered suggestive of neoplasm or other active disease on bone settings were also clinically important (type 3). If a lesion was judged as suspicious on original settings but characterized as degenerative change on bone settings the finding would be considered clinically important (type 3).
The average time required to use multiple settings was calculated by timing each radiologist in a random sample of 15 patients. The chi-square test was used to compare the data from the three categories of abnormality (type 1, type 2, and type 3).
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Overall, the conspicuity and characterization improved (type 2 or 3) in 81 (67%) of 121 abnormalities (p = 0.01). An improvement in lesion conspicuity or characterization classified as type 3 was present in 22 (18%) of 121 instances (p = 0.04). Of the 22 abnormalities rated type 3, 20 were overlooked on the original window and level settings. Two lesions (both bone abnormalities) were originally recognized as metastatic lesions but later identified as degenerative disease using bone settings.
Ratings were subdivided by type of study (chest or abdomen) because of differences in scanning technique and contrast use (differences that may have affected the characterization of liver abnormalities). For eight (44%) of 18 patients with liver abnormalities, chest CT displayed at liver windows showed an increase in conspicuity and characterization (type 2 or 3). For three patients (17%), CT showed improvement that substantially affected the final diagnosis (type 3). For 44 (80%) of 55 patients with liver abnormalities, abdominal CT displayed at liver windows showed an increase in conspicuity and characterization (type 2 or 3). For six patients (11%), CT showed improvement that substantially affected the final diagnosis (type 3). Nevertheless, the number of patients in these subsets was small and the improvements using liver windows were statistically insignificant. The average time required to use additional window and level settings was 40 ± 8 sec per patient.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In our study, radiologists benefited from using tissue-specific window settings for abdominal and chest CT. For example, the chest CT for a patient with bronchogenic carcinoma viewed with soft-tissue and lung windows did not show a large hepatic lesion on axial sections through the upper abdomen (Fig. 1A). The lesion was clearly evident and easily detected (type 3) using liver settings (Fig. 1B). Our institution's standard chest CT filming protocols did not previously include liver windows and this lesion might have gone undetected. In our study, the percentage of patients with additional findings of clinical importance (type 3) was higher on chest CT than on abdominal CT.
|
|
A second example illustrates the benefit of bone windows. On the abdominal CT for a patient with prostate carcinoma, a sclerotic vertebral body lesion probably representing a metastatic deposit was overlooked on soft-tissue windows (Fig. 2A). The lesion was easily detected on bone windows (Fig 2B).
|
|
In a film-based environment, printing multiple image sets results in increased film cost; increased expenditure of time to print, handle, sort, and hang film; and increased space requirements to display and store film. Moreover, multiple image sets cause increased navigational difficulty for radiologists. These factors limit the number of window and level combinations that are used in a film-based environment.
At our institution, the estimated cost per image-sheet including the film, chemicals, technologist time, and equipment is $2.68. Twelve images are printed on each sheet. Chest CT filmed at lung and soft-tissue settings requires three sheets for each sequence—a total of six sheets. The filming of bone windows would require an additional three sheets, and the filming of liver windows (for the upper abdomen) would require an additional one or two sheets—a total of up to 11 sheets. Printing the bone and liver window settings would cost an additional $10.72-13.40 per patient—a cost increase of 67-83%.
At our institution, typical abdominal CT is filmed with soft-tissue windows and lung windows for the lower thorax, requiring a total of six or seven sheets. Filming liver windows (for the upper abdomen) and bone windows would require three or four and four or five more sheets, respectively. Printing the additional window settings would cost an additional $18.76-24.12 per patient—a cost increase of 100-150%. It is difficult to quantify the additional cost in film-room sorting time and inefficiencies in hanging and interpreting the increased number of film sheets.
Soft-copy interpretation enables radiologists to view multiple settings without additional expense. In our study, an additional 40 sec (average) was required to view bone and liver windows for each patient's CT scans. Our study suggests that the added time is worth-while in exchange for the additional information obtained when using tissue-specific window and level settings.
In our study, the radiologists agreed that additional window and level settings offered direct and indirect diagnostic benefits. Indirect diagnostic benefits include the extra attention directed to the structures highlighted by specific settings. For example, because bones are poorly depicted on conventional settings, attention to bony structures may be minimal in clinical practice. However, when bone settings are applied, other structures are poorly depicted and the radiologist's attention is directed to the bones.
To our knowledge, only three studies specifically address the use of tailored CT window settings and each was performed in a film-based environment [6, 7, 8].
A recent paper by Mayo-Smith et al. [6] assessed the efficacy of liver settings. Their filmbased study examined abdominal CT using liver windows. The study revealed an increase in the number of lesions detected in 11.7% of abnormal livers when applying liver windows. The use of additional settings resulted in a diagnostic change for 1.7% of patients. The authors concluded that the use of additional settings was not cost-effective unless patients had a known malignancy or there was a high clinical index of suspicion.
Our study differs in that both liver and bone lesions were evaluated on chest and abdominal CT. We found greater benefit using liver windows than was reported in their study, and in particular, greater improvement in clinical importance. One major factor accounting for this difference may be the higher prevalence of disease in our study subjects at the Veterans Affairs hospital. In the study of Mayo-Smith et al. [6], 70% of livers were judged to be normal, whereas in our study, only 37% of livers were considered normal.
Another important difference is that the study of Mayo-Smith et al. [6] was conducted in a film-based environment, whereas in our study, images were evaluated using organ-specific windows on a soft-copy workstation. Although Mayo-Smith et al. concluded that the small clinical benefit revealed in their study might not be cost-effective in a film-based environment, the routine use of additional windows may be efficacious using an existing soft-copy system. Further benefits are derived when other organ-specific settings such as bone windows are used. Because clinical information is often lacking, we do not agree with the authors' recommendation that the use of additional windows be based on clinical suspicion.
Bach et al. [7] performed a retrospective study, also in a film-based environment, and concluded that supplemental bone windows are unnecessary for cancer patients. They reviewed the impression section of reports from 4683 body CT scans of cancer patients at an oncology specialty center to identify any description of a lesion suggestive of bone metastasis. All cases were filmed using supplemental bone windows. Bone windows were not considered worthwhile in 4160 patients because the impression section of reports did not mention a lesion. In an additional 271 patients, bone windows were considered unnecessary because a prior imaging study such as radiography, radionuclide bone scanning, MR imaging, or CT had already provided results indicative of a bone metastasis at the same site. In 206 patients, bone windows were considered noncontributory because there were extensive nonbone metastases, the bone finding was irrelevant by follow-up, or the bone finding was not pursued. In the remaining 46 patients, radiologists compared the bone and soft-tissue windows side by side in an unblinded fashion. In 45 of these patients, the lesions were equally visible on soft-tissue and bone windows. In summary, the study by Bach et al. identified only one patient of 4683 whose CT findings were clinically significant when viewed with additional window settings.
Our study and that of Bach et al. [7] differ in several respects. In their film-based study, the cost of printing bone window films was $15,814 ($1.20 per sheet). We incurred no additional expense using a soft-copy workstation. Their patient population was studied at an oncology referral center and was substantially different from ours. Because their patient population was derived from a specialized hospital, it is likely that a higher proportion of their patients had prior CT scans or ancillary studies than that of our patients. This factor would increase the number of patients for whom there was no clinical benefit of using bone windows because metastases were previously diagnosed. Bach et al. judged bone windows to be noncontributory in 89% of patients because no mention of a lesion suggestive of bone metastasis was present in the report. This definition does not allow for indeterminate findings on soft-tissue windows that could be confidently dismissed as insignificant on bone windows (e.g., sclerotic changes of degenerative disease). In the 46 patients for whom a bone finding was possibly important, all but one were considered equally visible on soft-tissue and bone windows. This result was in marked contrast to our reviewers who found 60% of bone lesions better depicted on bone settings and 27% of lesions clinically important. In part, we believe that this difference may be attributable to bias introduced by their inspection of patients in an unblinded fashion. Even if lesions were equally visible on soft-tissue and bone windows, it is doubtful that the skeletal structures would receive the same degree of scrutiny as occurs when bone settings are applied.
Vandemark et al. [8] evaluated the use of bone settings for the examination of bone metastases. These authors concluded that bone settings were useful in evaluating the bone metastases from breast cancer to chemotherapy. However, the authors compared the characterization at bone settings to radiography and radionuclide bone scanning rather than CT at soft-tissue settings. It is likely that they assumed that bone settings would be more useful than soft-tissue settings, an assumption supported by our data.
Our study had some notable limitations. The rating system we used was subjective; however, the definitions for each category were designed to minimize ambiguity, and representative training patients were shown before interpretation sessions to minimize variation. Nevertheless, the potential for subjectivity existed because a judgment of clinical importance was required.
Specific window and level settings may vary from institution to institution. This factor may limit the precise extrapolation of our results, although the general conclusions are likely applicable. In our study, we presented the radiologists with additional organ-specific window and level settings immediately after they had drawn conclusions from the conventional settings. This effort was intended to allow the radiologists to detect a true change in perception of abnormalities because of window and level differences. The presentation of images at a more delayed interval could introduce error caused by confounding factors such as simple day-to-day intraobserver variation.
In conclusion, we examined the use of additional window settings in the soft-copy environment. We found that soft-copy interpretation enables more effective evaluation of CT data sets. Reviewing organ- or tissue-specific window and level settings might increase diagnostic accuracy. Although the routine review of additional window and level settings might be impractical in the filmbased environment, it may be efficacious in the soft-copy setting.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  M. B. McCarville, H. M. Lederman, V. M. Santana, N. C. Daw, S. J. Shochat, C.-S. Li, and R. A. Kaufman Distinguishing Benign from Malignant Pulmonary Nodules with Helical Chest CT in Children with Malignant Solid Tumors. Radiology, May 1, 2006; 239(2): 514 - 520. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. I. Reiner, E. L. Siegel, and F. J. Hooper Accuracy of Interpretation of CT Scans: Comparing PACS Monitor Displays and Hard-Copy Images Am. J. Roentgenol., December 1, 2002; 179(6): 1407 - 1410. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. M. Patten, S. R. Gunberg, D. K. Brandenburger, and M. L. Richardson CT Detection of Hepatic and Splenic Injuries: Usefulness of Liver Window Settings Am. J. Roentgenol., October 1, 2000; 175(4): 1107 - 1110. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
