基于深度学习算法的胸部CT辅助诊断系统对急诊隐匿性肋骨骨折的检出效能研究  

Application of deep learning-based chest CT auxiliary diagnosis system in occult rib fractures patients

在线阅读下载全文

作  者:周柳含 周颖俊 Zhou Liuhan;Zhou Yingjun(Department of Radiology,Xiangtan Central Hospital,Xiangtan 411100,China)

机构地区:[1]湘潭市中心医院放射科,湘潭411100

出  处:《现代仪器与医疗》2024年第6期76-80,共5页Modern Instruments & Medical Treatment

摘  要:目的评估基于深度学习算法(Deep Learning,DL)的胸部CT辅助诊断系统对肋骨隐匿性骨折的诊断价值。方法在本回顾性研究中,收集湘潭市中心医院2019年1月—2023年6月95名疑似肋骨骨折行急诊CT检查患者的影像学资料,导入DL辅助诊断系统,并获取其影像医师原始报告结果。以患者2~3周胸部CT复查结果为确诊标准,对比DL辅助诊断系统与影像医师对肋骨骨折的诊断效能。结果95例患者中,经复查确诊肋骨骨折84例。DL辅助诊断系统诊断肋骨骨折的敏感度、特异度、阳性预测值以及阴性预测值分别为98.81%、81.82%、97.64%以及90.00%。影像医师诊断肋骨骨折的敏感度、特异度、阳性预测值以及阴性预测值分别为80.95%、81.82%、97.14%以及36.00%。DL辅助诊断系统诊断肋骨的敏感度、阴性预测值显著高于影像医师(Z=4.248,P<0.001)。结论基于DL辅助诊断系统能够自动识别肋骨骨折,且漏诊率较之影像医师显著降低,对优化急诊肋骨骨折诊疗过程有重要的作用。Objective To evaluate the diagnostic efficacy and potential application value of deep learning-based chest CT auxiliary diagnosis system in occult rib fractures patients.Methods In this retrospective study,95 patients with suspected rib fracture for emergency CT examination,who were randomly selected,and then introduced them into DL auxiliary diagnosis system and obtained the original report results of Radiologist.Based on the results of chest CT Reexamination in 2-3 weeks,the diagnostic efficiency of DL auxiliary diagnosis system and Radiologist was compared.Results among 95 patients,84 cases were diagnosed as rib fracture by reexamination.The sensitivity,specificity,positive predictive value and negative predictive value of DL auxiliary diagnostic system for rib fracture were 98.81%,81.82%,97.64%and 90.00%,respectively.The sensitivity,specificity,positive predictive value and negative predictive value of Radiologist were 80.95%,81.82%,97.14%and 36.00%respectively.The sensitivity and negative predictive value of DL auxiliary diagnostic system in the diagnosis of rib were significantly higher than those of radiologists(Z=4.248,P<0.001).Conclusion DL auxiliary diagnosis system can automatically identify rib fracture with low missed diagnosis rate,which plays an important role in optimizing the diagnosis and treatment process of emergency rib fracture.

关 键 词:CT 深度学习 影像医师 肋骨骨折 诊断效能 

分 类 号:R197.39[医药卫生—卫生事业管理] TH77[医药卫生—公共卫生与预防医学]

 

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