基于深度学习的人工智能胸部CT肺结节检测效能评估

被引:119
作者
李欣菱 [1 ]
郭芳芳 [2 ]
周振 [3 ]
张番栋 [3 ]
王卿 [1 ]
彭志君 [1 ]
苏大同 [1 ]
范亚光 [4 ]
王颖 [1 ]
机构
[1] 天津医科大学总医院放射科
[2] 新乡医学院第一附属医院放射科
[3] Deepwise Healthcare
[4] 天津医科大学总医院,天津市肺癌研究所,天津市肺癌转移与肿瘤微环境重点实验室
关键词
计算机体层成像; 肺结节; 深度学习; 人工智能; 检出;
D O I
暂无
中图分类号
TP18 [人工智能理论]; R816.4 [胸部及呼吸系];
学科分类号
100106 [放射医学]; 140502 [人工智能];
摘要
背景与目的肺结节精确检测是实现肺癌早诊的基础。基于深度学习的人工智能在肺内结节检测领域发展迅速,对其效能进行验证是促进其应用于临床的前提。本研究旨在评估基于深度学习技术的人工智能软件在胸部计算机断层扫描(computed tomography, CT)恶性及非钙化结节检出中的价值。方法由天津医科大学总医院自建胸部CT肺结节数据库中随机抽取200例胸部CT数据,包含病理证实的肺癌及随访结节病例,导入肺结节人工智能识别系统,记录软件自动识别结节,并与原始影像报告结果进行对比。人工智能软件及阅片者检测到的结节由2名胸部专家进行评估并记录其大小及特征。计算灵敏度、假阳性率评估人工智能软件及医师的结节检测效能,应用McNemar检验确定二者之间是否存在显著性差异。结果 200例胸部多层螺旋CT共包含非钙化结节889枚,其中肺癌结节133枚,小于5 mm结节442枚。人工智能及放射科医师肺癌检出率皆为100%。人工智能软件结节检测灵敏度明显高于放射科医师(99.1%vs 43%, P<0.001)。人工智能总体假阳性率为每例CT 4.9个,排除5 mm以下结节后降为1.5个。结论基于深度学习的人工智能软件能实现恶性肺结节的无漏诊检出,具有较医师更高的结节检出灵敏度,在排除微小结节后可降低假阳性率。
引用
收藏
页码:336 / 340
页数:5
相关论文
共 7 条
[1]
Validation; comparison; and combination of algorithms for automatic detection of pulmonary nodules in computed tomography images: The LUNA16 challenge.[J].Arnaud Arindra Adiyoso Setio;Alberto Traverso;Thomas de Bel;Moira S.N. Berens;Cas van den Bogaard;Piergiorgio Cerello;Hao Chen;Qi Dou;Maria Evelina Fantacci;Bram Geurts;Robbert van der Gugten;Pheng Ann Heng;Bart Jansen;Michael M.J. de Kaste;Valentin Kotov;Jack Yu-Hung Lin;Jeroen T.M.C. Manders;Alexander Sóñora-Mengana;Juan Carlos García-Naranj
[2]
Global Cancer Statistics, 2012 [J].
Torre, Lindsey A. ;
Bray, Freddie ;
Siegel, Rebecca L. ;
Ferlay, Jacques ;
Lortet-Tieulent, Joannie ;
Jemal, Ahmedin .
CA-A CANCER JOURNAL FOR CLINICIANS, 2015, 65 (02) :87-108
[3]
Automated Identification of Dementia Using FDG-PET Imaging.[J].Yong Xia;Shen Lu;Lingfeng Wen;Stefan Eberl;Michael Fulham;David Dagan Feng;Annalena Venneri.BioMed Research International.2014,
[4]
Prognostic significance of using solid versus whole tumor size on high-resolution computed tomography for predicting pathologic malignant grade of tumors in clinical stage IA lung adenocarcinoma: A multicenter study [J].
Tsutani, Yasuhiro ;
Miyata, Yoshihiro ;
Nakayama, Haruhiko ;
Okumura, Sakae ;
Adachi, Shuji ;
Yoshimura, Masahiro ;
Okada, Morihito .
JOURNAL OF THORACIC AND CARDIOVASCULAR SURGERY, 2012, 143 (03) :607-612
[5]
Effect of CAD on Radiologists' Detection of Lung Nodules on Thoracic CT Scans: Analysis of an Observer Performance Study by Nodule Size [J].
Sahiner, Berkman ;
Chan, Heang-Ping ;
Hadjiiski, Lubomir M. ;
Cascade, Philip N. ;
Kazerooni, Ella A. ;
Chughtai, Aamer R. ;
Poopat, Chad ;
Song, Thomas ;
Frank, Luba ;
Stojanovska, Jadranka ;
Attili, Anil .
ACADEMIC RADIOLOGY, 2009, 16 (12) :1518-1530
[6]
Commercially Available Computer-Aided Detection System for Pulmonary Nodules on Thin-Section Images Using 64 Detectors-Row CT.[J].Masahiro Yanagawa;Osamu Honda;Shigeyuki Yoshida;Yusuke Ono;Atsuo Inoue;Tadahisa Daimon;Hiromitsu Sumikawa;Naoki Mihara;Takeshi Johkoh;Noriyuki Tomiyama;Hironobu Nakamura.Academic Radiology.2009, 8
[7]
Evaluation of a computer aided detection system for lung nodules with ground glass opacity component on multidetector-row CT..Beigelman-Aubry C;Hill C;Boulanger X;Brun Al;Leclercq D;Golmard Jl;Grenier Pa;Lucidarme O;.Journal de radiologie.2009,