摘要
目的:探讨正电子发射断层扫描(PET)/计算机断层扫描(CT)结合人工智能与512层螺旋CT对肺内高凝聚结节定性分析价值。方法:回顾性分析220例肺内结节患者的检查资料,对研究对象均接受PET/CT检查与螺旋CT检查,针对PET/CT检查影像采用人工智能分析。以病理结果为金标准,对比不同方法诊断疾病的价值,比较阅片时间与图像质量,计算不同方法的诊断效能。结果:病理证实175例良性结节,45例恶性结节。PET/CT结合人工智能分析与病理结果一致性为0.7564,高于螺旋CT与病理结果一致性0.512(P<0.05)。PET/CT结合人工智能分析对炎性假瘤检出率99.02%高于螺旋CT检出率93.14%(P<0.05)。PET/CT结合人工智能分析诊断灵敏度为95.56%,高于螺旋CT检查的77.78%(P<0.05)。结论:采取PET/CT结合人工智能分析对肺内高浓聚结节定性分析价值高,可缩短阅片时间,提高诊断的灵敏度。
Objective:To explore the qualitative analysis value of positron emission tomography(PET)/computed tomography(CT)combined with artificial intelligence and 512 slice spiral CT for high coagulation nodules in the lungs.Methods:A retrospective analysis was conducted on the examination data of 220 patients with pulmonary nodules.All subjects underwent PET/CT and spiral CT examinations,and artificial intelligence analysis was used for PET/CT imaging.Using pathological results as the gold standard,compare the diagnostic value of different methods for diseases,compare the viewing time and image quality,and calculate the diagnostic efficacy of different methods.Results:Pathology confirmed 175 cases of benign nodules and 45 cases of malignant nodules.The consistency between PET/CT combined with artificial intelligence analysis and pathological results was 0.7564,which was higher than the consistency between spiral CT and pathological results of 0.512(P<0.05).The detection rate of inflammatory pseudotumors using PET/CT combined with artificial intelligence analysis was 99.02%,which was higher than the detection rate of 93.14%using spiral CT(P<0.05).The diagnostic sensitivity of PET/CT combined with artificial intelligence analysis was 95.56%,which was higher than the 77.78%of spiral CT examination(P<0.05).Conclusion:The combination of PET/CT and artificial intelligence analysis has high value in qualitative analysis of high concentration nodules in the lungs,which can shorten the time of film reading and improve the sensitivity of diagnosis.
作者
郭佳
林莉
廖娟彬
黄谋清
曾小建
廖忠剑
GUO Jia;LIN Li;LIAO Juan-bin;HUANG Mou-qing;ZENG Xiao-jian;LIAO Zhong-jian(Nuclear Medicine Department,Ganzhou People’s Hospital,Jiangxi Ganzhou 341000)
出处
《中国医疗器械信息》
2024年第18期13-15,19,共4页
China Medical Device Information
基金
赣州市卫生健康委员会市级科研计划项目(项目名称:PET/CT结合人工智能与512层螺旋CT对肺内高浓聚结节定性分析临床研究,项目编号:2020-2-11)