影像诊断工作站显示器的选择对孤立性肺结节识读影响的研究

Influences of resolution chosen for image diagnostic monitors on detection accuracy of single pulmonary nodules

目的通过对胸部DR影像上孤立性肺结节检出准确率的比较来评价PACS影像诊断工作站上不同分辨率的医用显示器和普通显示器的诊断效能。方法选取经CT证实的孤立性肺结节100例,多发肺结节24例和疑似病例15例(直径<30mm,两肺数目不超过3个并散在分布)。2位高年资医生和3位低年资医生通过3种显示器(BarcoMGD521MKII(2560×2048,75Hz)医用RCT,EIZORadiForceG20(1200×1600,60Hz)医用LCD和DellRoundRockTexas78682(1280×1024,85Hz)普通RCT)集中进行3次阅片,采用5分法:肯定有、可能有、不确定、可能没有、肯定没有来判断胸片上是否有肺结节及结节个数。采用SPSS13.0软件中受试者工作特征(ROC)曲线法进行统计分析。结果①高年资医师使用BarcoRCT的Az值(Az=0.770)明显高于EIZOLCD(Az=0.724)(P<0.05),而在普通(Az=0.714)和医用显示器之间无明显差别(P>0.05)。低年资医师使用EIZOLCD的Az值(Az=0.640)明显高于DellRCT(Az=0.598)(P<0.05),在两种医用显示器之间则存在小的但具有统计学显著性意义的差别(P<0.05)。②在各种情况下阅片,高年资医师的诊断准确率都明显高于低年资医师(P<0.05)。高年资医师使用普通显示器的Az值较低年资医师使用两种医用显示器时都要高(P<0.05)。结论①出于成本效益的考虑,科室可以安排低年资医师使用较低配置医用显示器完成初步诊断,高年资医师应按标准使用医用显示器(2560×2048)以保证最终诊断结果的可靠性。②对于普通放射诊断科室,完全可以采用医用显示器与普通显示器相结合组成诊断工作站系统。不同分辨率显示器的合理配置,可获得较好的性价比。

Objective The purposes of this study is to assess the influence and diagnostic value of professional image diagnostic monitors with different resolution and commercially available monitors used in PACS by determining the rate of detection of single pulmonary nodules （SPNs） on soft copy images of chest direct digital radiography. Methods One hundred chest radiographs showing a solitary pulmonary nodule, 24 patients with multi-pulmonary nodules and 15 cases without pulmonary nodule were selected for this study （all smaller than 30 mm in diameter and not more than three dispersing on two lung fields）, and CT scans of the same patient served as the gold standard for the presence of pulmonary nodules. Two experienced radiologists and three residents three interpreted 139 chest images by using diagnostic workstation monitors of PACS in three brands： Barco MGD521 MKII （2560 × 2048 pixels, 75 Hz） professional RCT, EIZO RadiForce G20 （1200 × 1600 pixels, 60 Hz） professional LCD and Dell Round Rock Texas78682 （1280 × 1024 pixels, 85 Hz） commercially available RCT and marked their confidence of the nodules. The presence or absence of SPNs was investigated with a five-point confidences rating scale： ①definitely positive, ② probably positive, ③ indeterminate, ④probably negative,⑤ definitely negative. The interpreting results were analyzed by SPSS 13. 0 software-the receiver operating characteristic （ROC） curve. Results ①To experienced radiologists, the average Az values in Barco RCT （Az= 0. 770） were obviously superior to that of EIZO LCD （Az=0. 724） （P〈0. 05）, but there was no significant difference between professional and commercially available monitors （Az=0. 714） （P〉0.05）. For the residents, the difference of observer performance in three monitors were conspicuous （P 〈0.05）. ②In all monitors, the observer performances of experienced radiologists were higher than that of residents （P〈 0. 05）. The Az values for experienced radiologists were higher on commercially available monitors than that for residents on professional monitors （P〈0.05）. Conclusion ①Proceeding from the consideration of the cost efficiency, the administrative offices may arrange less experienced radiologists to use professional monitors with less resolution for primary diagnostic tasks, the experienced radiologists should better than using professional monitors（2560 × 2048） by to guarantee finally diagnoses reliability. ②For general radiodiagnosis department, it is definitely to combine the diagnosis workstation system with commercially available monitors and professional monitors completely. The resolution of imaging diagnostic monitors can af- fect diagnostic performance and reasonable equipment for the detection of subtle abnormality may result in better cost-efficacy and diagnostic accuracy.