Study on the consistency between artificial intelligence and radiologist in interpreting CT signs of solid lung masses

doi:10.3877/cma.j.issn.2095-2015.2022.06.005

Abstract

Abstract:

Objective

To analyze the consistency between artificial intelligence and radiologists in interpreting CT signs of solid lung masses, and explore the value of artificial intelligence in clinical application.

Methods

A retrospective analysis was performed on 58 cases of solid lung masses confirmed by pathology in Huai ′an Hospital Affiliated to Xuzhou Medical University from January 2020 to January 2022.CT thin layer data package was reconstructed and the tumor signs(including lobulation sign, burr sign, vacuole sign, vascular cluster sign and pleural depression sign)were analyzed by artificial intelligence software.At the same time, two radiologists analyzed the above characteristics of the tumors respectively.Consistency test was used to analyze the consistency of interpretation results by radiologists and artificial intelligence.

Results

The consistency of the radiologist′s two interpretation of whether the tumors had lobulation sign, burr sign, vacuole sign, vascular convergence sign, pleural depression sign and bronchial inflation sign was excellent(Kappa value ranged from 0.862 to 0.965, all P<0.001), and the consistency of the two radiologists′ interpretation results was excellent(Kappa value ranged from 0.826 to 0.928, all P<0.001). The consistency of CT signs of tumors interpreted by radiologists and artificial intelligence was excellent(Kappa value ranged from 0.770 to 0.906, all P<0.001), among which the consistency of lobulation sign was the highest(Kappa=0.906), and the consistency of vascular bundle sign was the lowest(Kappa=0.770).

Conclusion

Artificial intelligence can reliably evaluate the CT signs of solid lung masses, and improve the efficiency of radiologists.

Key words: Artificial intelligence, Lung, Tumor, Tomography, X-ray computer

Lei Han, Xiaoyu Chen, Nannan Wang. Study on the consistency between artificial intelligence and radiologist in interpreting CT signs of solid lung masses[J]. Chinese Journal of Digestion and Medical Imageology(Electronic Edition), 2022, 12(06): 348-350.

/ / Recommend

Add to citation manager EndNote|Ris|BibTeX

URL: https://zhxhbyyxzz.cma-cmc.com.cn/EN/10.3877/cma.j.issn.2095-2015.2022.06.005

https://zhxhbyyxzz.cma-cmc.com.cn/EN/Y2022/V12/I06/348

Figures/Tables 4

References 20

1	Deshpand R，Chandra M，Rauthan A．Evolving trends in lung cancer：Epidemiology，diagnosis，and management[J]．Indian J Cancer，2022，59(Supplement)：S90-S105.
2	Zhang J，Ijzerman MJ，Oberoi J，et al.Time to diagnosis and treatment of lung cancer：A systematic overview of risk factors，interventions and impact on patient outcomes[J]．Lung Cancer，2022，166：27-39.
3	舒静，张涵，赵振国，等．肺炎性假瘤、周围型肺癌CT征象特征及其鉴别诊断价值[J]．中国CT和MRI杂志，2021，19(5)：35-37.
4	Tao G，Zhu L，Chen Q，et al.Prediction of future imagery of lung nodule as growth modeling with follow-up computed tomography scans using deep learning：a retrospective cohort study[J]．Transl Lung Cancer Res，2022，11(2)：250-262.
5	梁雪，潘金彬，丁建民，等．超声造影LI-RADS与增强CT/MRI LI-RADS对肝微小病灶分类的一致性及差异性研究[J]．中华超声影像学杂志，2021，30(11)：938-943.
6	温艳，于连政，杜灵彬，等．中国3省城市癌症早诊早治项目地区肺癌高危人群的低剂量螺旋CT筛查依从性及相关因素分析[J]．中华预防医学杂志，2021，55(5)：633-639.
7	孟瑞瑞，刘圆圆，青浩渺，等．肺癌高危人群低剂量螺旋CT筛查的临床分析[J]．放射学实践，2021，36(1)：71-75.
8	Gao Y，Hua M，Lv J，et al.Reproducibility of radiomic features of pulmonary nodules between low-dose CT and conventional-dose CT[J]．Quant Imaging Med Surg，2022，12(4)：2368-2377.
9	Lan CC，Hsieh MS，Hsiao JK，et al.Deep Learning-based Artificial Intelligence Improves Accuracy of Error-prone Lung Nodules[J]．Int J Med Sci，2022，19(3)：490-498.
10	段秀杰，李福元，付玉存．周围型非小细胞肺癌CT征象与临床病理分型的关系[J]．现代肿瘤医学，2020，28(15)：2622-2626.
11	Wei S，Shi B，Zhang J，et al.Differentiating mass-like tuberculosis from lung cancer based on radiomics and CT features[J]Transl Cancer Res，2021，10(10)：4454-4463.
12	Manickavasagam R，Selvan S，Selvan Mary.CAD system for lung nodule detection using deep learning with CNN[J]．Med Biol Eng Comput，2022，60(1)：221-228.
13	Potter AL，Rosenstein AL，Kiang MV，et al.Association of computed tomography screening with lung cancer stage shift and survival in the United States：quasi-experimental study[J]．BMJ，2022，376：e069008.
14	Han X，Luo N，Xu L，et al.Artificial intelligence stenosis diagnosis in coronary CTA：effect on the performance and consistency of readers with less cardiovascular experience[J]．BMC Med Imaging，2022，22(1)：28.
15	Lyu SY，Zhang Y，Zhang MW，et al.Diagnostic value of artificial intelligence automatic detection systems for breast BI-RADS 4 nodules[J]．World J Clin Cases，2022，10(2)：518-527.
16	Deng J，Zhao M，Li Q，et al.Implementation of artificial intelligence in the histological assessment of pulmonary subsolid nodules[J]．Transl Lung Cancer Res，2021，10(2)：4574-4586.
17	Zhang H，Meng D，Cai S，et al.The application of artificial intelligence in lung cancer：a narrative review[J]．Transl Cancer Res，2021，10(5)：2478-2487.
18	Huang G，Wei X，Tang H，et al.A systematic review and meta-analysis of diagnostic performance and physicians′ perceptions of artificial intelligence(AI)-assisted CT diagnostic technology for the classification of pulmonary nodules[J]．J Thorac Dis，2021，13(8)：4797-4811.
19	Lancaster HL，Zheng S，Aleshina OO，et al.Outstanding negative prediction performance of solid pulmonary nodule volume AI for ultra-LDCT baseline lung cancer screening risk stratification[J]．Lung Cancer，2022，165：133-140.
20	于广浩，李为民，高杨，等．人工智能系统在CT肺小结节筛查中的准确率及检出时间分析[J]．中国医药科学，2021，11(21)：193-195，208.

CT特征	分叶征	毛刺征	空泡征	血管集束征	胸膜凹陷征	支气管充气征
Kappa值	0.931	0.922	0.896	0.862	0.930	0.965
P值	<0.001	<0.001	<0.001	<0.001	<0.001	<0.001

CT特征	分叶征	毛刺征	空泡征	血管集束征	胸膜凹陷征	支气管充气征
Kappa值	0.931	0.922	0.896	0.862	0.930	0.965
P值	<0.001	<0.001	<0.001	<0.001	<0.001	<0.001

CT特征	分叶征	毛刺征	空泡征	血管集束征	胸膜凹陷征	支气管充气征
Kappa值	0.894	0.847	0.902	0.826	0.856	0.928
P值	<0.001	<0.001	<0.001	<0.001	<0.001	<0.001

CT特征	分叶征	毛刺征	空泡征	血管集束征	胸膜凹陷征	支气管充气征
Kappa值	0.894	0.847	0.902	0.826	0.856	0.928
P值	<0.001	<0.001	<0.001	<0.001	<0.001	<0.001

CT特征	分叶征	毛刺征	空泡征	血管集束征	胸膜凹陷征	支气管充气征
Kappa值	0.906	0.802	0.847	0.770	0.855	0.811
P值	<0.001	<0.001	<0.001	<0.001	<0.001	<0.001

Please choose a citation manager

Content to export