计算机辅助检测系统在数字化胸片肺结节检出中的应用价值  被引量:8

Feasibility of computer-aided nodule detection system in digital chest radiographs

在线阅读下载全文

作  者:徐岩[1] 马大庆[1] 贺文[1] 

机构地区:[1]首都医科大学附属北京友谊医院放射科,100050

出  处:《中华放射学杂志》2007年第11期1169-1173,共5页Chinese Journal of Radiology

基  金:北京市自然科学基金(7062020)

摘  要:目的评价计算机辅助检测(computer-aided detection,CAD)系统在数字化胸片肺结节检出中的应用价值及其对放射科医师的辅助作用。方法选取数字化胸片328例。由2名专家组医师应用 IQQAT^(TM)-Chest 系统阅读具有结节样阴影的胸片,2人意见达成一致后标记结节的位置和大小并保存标记结果,将标记结果作为金标准来评估 CAD 系统检测肺结节的能力。由8名不同年资的放射科医师首先独立阅读具有结节样阴影的数字 X 线摄影(DR)胸片并保存诊断结果,然后再应用CAD 系统阅读胸片,将最终结果存入 CAD 系统。应用受试者操作特征曲线(ROC)和配对 t 检验来分析放射科医师应用 CAD 系统前后在肺结节检测能力上的差异。结果在100例 DR 胸片中,金标准结节151个,CAD 系统肺结节检测敏感性为78.1%(118.0/151),低年资放射科医师不用和应用 CAD系统时,肺结节的检测敏感性分别为62.4%(94.2/151)和77.4%(116.8/151),ROC 曲线下面积分别为0.769 和0.836,二者之间的差异具有统计学意义(P<0.01);高年资放射科医师不用和应用CAD 系统时,肺结节的检测敏感性分别为73.8%(111.5/151)和76.2%(115.0/151),ROC 曲线下面积分别为0.820和0.827,二者之间的差异无统计学意义(P>0.05)。结论 CAD 系统能够辅助放射科医师提高肺小结节的检测敏感性,对低年资医师的帮助更大。Objective To assess the usefulness of a computer-aided detection (CAD) system on detecting the pulmonary nodules in digital chest radiography. Methods Three hundred and twenty-eight digital chest radiographies along with corresponding chest computed tomography were selected for this study. Two senior chest radiologists interpreted these cases using the CAD system and marked the locations and sizes of all nodules with consensus, which were stored in a computer system as Gold Standard to evaluate the performance of the CAD system. Eight radiologists of various experience read these selected cases without and with the aid of CAD system and their results were stored in a computer system. The radiologists' performance was evaluated by using the receiver operating characteristic analysis. Comparison is made regarding the difference in Az values when CAD was not used versus when CAD was used by using the paired-samples t test. Results For small nodules, the sensitivity of this computer system used by the two senior chest radiologists in 100 digital radiographs was 78.1% (118.0/151). Without and with the assistance of CAD system, the nodule detection sensitivity of the radiology residents was 62.4% (94. 2/151 ) ,77.4% ( 116. 8/151 ) respectively, and the area under the curve of radiology residents was 0. 769, 0. 836 respectively. Statistical analysis showed there was statistically significant difference between interpretation without and with the CAD system ( P 〈 0. 05 ). The nodule detection sensitivity of the attending radiologists was 73.8% ( 111.5/151 ) ,76. 2% ( 115. 0/151 ) respectively with and without the assistance of CAD system, and the area under the curve of the attending radiologists increased from 0. 820 to 0. 826 after using the CAD system, which had no statistical significance ( P 〉 0. 05 ). Conclusions This CAD system can help to enhance the radiologists' ability to detect small pulmonary nodules. This is especially helpful for radiology residents.

关 键 词:硬币病变  X线摄影术 图像处理 计算机辅助 

分 类 号:R816.4[医药卫生—放射医学]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象