人工智能联合超微血管成像技术在乳腺结节诊断中的价值

丛小宇; 笪应芬; 汪成; 梁新凤; 陈民; 袁杰

doi:10.7619/jcmp.20230769

人工智能联合超微血管成像技术在乳腺结节诊断中的价值

丛小宇¹,
笪应芬^1, ,,
汪成²,
梁新凤¹,
陈民¹,
袁杰¹

1.
上海交通大学医学院附属第九人民医院黄浦分院超声科, 上海, 200011
2.
上海交通大学医学院附属第九人民医院黄浦分院乳腺外科, 上海, 200011

基金项目:

上海市科技计划项目 20ZR1432400

详细信息

通讯作者:
笪应芬, E-mail: dayingfen@126.com

中图分类号: R445.1;R737.9
计量
- 文章访问数: 163
- HTML全文浏览量: 37
- PDF下载量: 19
出版历程
- 收稿日期: 2023-03-12
- 修回日期: 2023-05-11
- 网络出版日期: 2023-09-05
- 刊出日期: 2023-08-27

Value of artificial intelligence combined with super microvascular imaging technology in the diagnosis of breast nodules

1.
Department of Ultrasound, Huangpu Branch of Affiliated Ninth People's Hospital of School of Medicine of Shanghai Jiao Tong University, Shanghai, 200011
2.
Department of Breast Surgery, Huangpu Branch of Affiliated Ninth People's Hospital of School of Medicine of Shanghai Jiao Tong University, Shanghai, 200011

摘要

摘要:
目的
探讨人工智能S-Detect技术联合智能三维超微血管成像(3D-SMI)技术对乳腺结节良恶性的诊断价值。
方法
选取2021年1月—2023年2月151例(192个结节)乳腺病变患者为研究对象。采用常规超声检查、S-Detect技术、智能3D-SMI技术对乳腺结节进行良恶性鉴别, 以术后病理结果为金标准，绘制受试者工作特征(ROC)曲线，分析常规超声检查、S-Detect技术、智能3D-SMI技术及三者联合诊断对乳腺结节良恶性的诊断效能。
结果
192个结节中，病理证实良性结节112个，恶性结节80个。常规超声检查、S-Detect技术、智能3D-SMI技术及三者联合诊断的敏感度、特异度和准确度分别为70.00%、83.93%、78.13%, 78.75%、79.46%、79.17%, 71.25%、93.75%、84.38%, 90.00%、80.36%、84.38%。三者联合诊断的诊断效能较常规超声检查、S-Detect技术高，差异有统计学意义(Z=2.567, P=0.010; Z=2.533, P=0.011)。常规超声检查、S-Detect技术、智能3D-SMI技术的曲线下面积(AUC)比较，差异无统计学意义(P>0.05)。
结论
人工智能S-Detect技术与智能3D-SMI技术联合应用可辅助常规超声诊断乳腺结节的良恶性，有助于提高诊断的准确率。
- S-Detect技术 /
- 超微血管成像 /
- 乳腺肿瘤 /
- 超声检查
Abstract:
Objective
To explore the diagnostic value of artificial intelligence S-Detect technique and smart three-dimensional super microvascular imaging (3D-SMI) technique for diagnosis of benign and malignant breast nodules.
Methods
A total of 151 patients with breast lesions (192 nodules) in our hospital from January 2021 to February 2023 were selected as study objects. Conventional ultrasound examination, S-Detect technique, smart 3D-SMI technique were used to identify benign and malignant breast nodules, with the postoperative pathological results as the gold standard, the receiver operating characteristic(ROC) curve was drawn, the diagnostic efficacy of conventional ultrasonography, S-Detect technology, smart 3D-SMI technique and the combined diagnosis of three techniques in the differentiation of benign and malignant breast nodules was compared.
Results
Of 192 nodules, 112 nodules were diagnosed as benign ones and 80 as malignant ones. The sensitivity, specificity and accuracy of conventional ultrasound examination, S-Detect technique, smart 3D-SMI technique and the combination were 70.00%, 83.93%, 78.13%; 78.75%, 79.46%, 79.17%; 71.25%, 93.75% and 84.38%; 90%, 80.36% and 84.38%, respectively. The diagnostic efficacy of the their combined diagnosis was higher than that of conventional ultrasound examination and S-Detect technology (Z=2.567, P=0.010; Z=2.533, P=0.011). There was no significant difference in area under the curve(AUC) among conventional ultrasound examination, S-Detect technique and smart 3D-SMI technique (P>0.05).
Conclusion
The combined application of artificial intelligence S-Detect technique and smart 3D-SMI technique can assist the conventional ultrasound in the diagnosis of benign and malignant breast nodules, and help to improve the diagnostic accuracy.
- S-Detect technique /
- super microvascular imaging /
- breast tumor /
- ultrasonography

HTML全文

图 1 常规超声检查、S-Detect技术、智能3D-SMI技术和三者联合诊断的ROC曲线

下载: 全尺寸图片幻灯片

表 1 不同检查方法诊断良恶性结节的结果例

方法	结果	病理结果		合计
方法	结果	恶性(n=80)	良性(n=112)	合计
常规超声检查	恶性	56	18	74
	良性	24	94	118
S-Detect技术	恶性	63	23	86
	良性	17	89	106
智能3D-SMI技术	恶性	57	7	64
	良性	23	105	128
三者联合诊断	恶性	72	22	94
	良性	8	90	98

下载: 导出CSV

表 2 不同检查方法对乳腺良恶性结节的诊断效能

方法	敏感度/%	特异度/%	准确度/%	阳性预测值/%	阴性预测值/%
常规超声检查	70.00	83.93	78.13	75.68	79.66
S-Detect技术	78.75	79.46	79.17	73.26	83.96
智能3D-SMI技术	71.25	93.75	84.38	89.06	82.03
三者联合诊断	90.00	80.36	84.38	76.60	91.84

下载: 导出CSV

参考文献(19)

[1]	何思怡, 李贺, 曹毛毛, 等. 全球及我国女性乳腺癌疾病负担年龄分布及变化趋势[J]. 中国肿瘤, 2023, 32(1): 1-7. https://www.cnki.com.cn/Article/CJFDTOTAL-ZHLU202301001.htm
[2]	ALAREF A, HASSAN A, SHARMA KANDEL R, et al. Magnetic resonance imaging features in different types of invasive breast cancer: a systematic review of the literature[J]. Cureus, 2021, 13(3): e13854.
[3]	LIU M J, MEN Y M, ZHANG Y L, et al. Improvement of diagnostic efficiency in distinguishing the benign and malignant thyroid nodules via conventional ultrasound combined with ultrasound contrast and elastography[J]. Oncol Lett, 2017, 14(1): 867-871. doi: 10.3892/ol.2017.6245
[4]	STROBEL K, SCHRADING S, HANSEN N L, et al. Assessment of BI-RADS category 4 lesions detected with screening mammography and screening US: utility of MR imaging[J]. Radiology, 2015, 274(2): 343-351. doi: 10.1148/radiol.14140645
[5]	戚瑞祥, 朱罗茜, 方建华, 等. S-Detect技术在乳腺非肿块型病变中的诊断价值[J]. 中国超声医学杂志, 2021, 37(3): 251-255. doi: 10.3969/j.issn.1002-0101.2021.03.004
[6]	ZHANG X Y, ZHANG L, LI N, et al. Vascular index measured by smart 3-D superb microvascular imaging can help to differentiate malignant and benign breast lesion[J]. Cancer Manag Res, 2019, 11: 5481-5487. doi: 10.2147/CMAR.S203376
[7]	张曼琪, 查海玲, 刘心培, 等. 超声检查评估乳腺癌新辅助化疗疗效的准确性及影响因素分析[J]. 实用临床医药杂志, 2022, 26(13): 45-49, 54. doi: 10.7619/jcmp.20220387
[8]	薛雯, 杨柳茵, 范丽, 等. 超微血管成像技术鉴别乳腺良恶性病变[J]. 中国医学影像技术, 2019, 35(1): 77-81. https://www.cnki.com.cn/Article/CJFDTOTAL-ZYXX201901026.htm
[9]	GUO R R, LU G L, QIN B J, et al. Ultrasound imaging technologies for breast cancer detection and management: a review[J]. Ultrasound Med Biol, 2018, 44(1): 37-70. doi: 10.1016/j.ultrasmedbio.2017.09.012
[10]	LY A, ONO J C, HUGHES K S, et al. Fine-needle aspiration biopsy of palpable breast masses: patterns of clinical use and patient experience[J]. J Natl Compr Canc Netw, 2016, 14(5): 527-536. doi: 10.6004/jnccn.2016.0061
[11]	BARTOLOTTA T V, ORLANDO A A M, SPATAFORA L, et al. S-Detect characterization of focal breast lesions according to the US BI RADS lexicon: a pictorial essay[J]. J Ultrasound, 2020, 23(2): 207-215. doi: 10.1007/s40477-020-00447-w
[12]	邢博缘, 董梦, 刘冬婷, 等. 超声S-Detect技术对直径≤2 cm乳腺小结节的诊断价值[J]. 中国超声医学杂志, 2022, 38(6): 626-629. doi: 10.3969/j.issn.1002-0101.2022.06.009
[13]	ZHAO C Y, XIAO M S, LIU H, et al. Reducing the number of unnecessary biopsies of US-BI-RADS 4a lesions through a deep learning method for residents-in-training: a cross-sectional study[J]. BMJ Open, 2020, 10(6): e035757. doi: 10.1136/bmjopen-2019-035757
[14]	曹力, 张会萍, 周毓青, 等. 不同年资超声科医师应用S-Detect技术诊断乳腺结节的价值研究[J]. 肿瘤影像学, 2022, 31(4): 439-443. https://www.cnki.com.cn/Article/CJFDTOTAL-YXYX202204013.htm
[15]	FUJIOKA T, KUBOTA K, MORI M, et al. Distinction between benign and malignant breast masses at breast ultrasound using deep learning method with convolutional neural network[J]. Jpn J Radiol, 2019, 37(6): 466-472. doi: 10.1007/s11604-019-00831-5
[16]	李伟伟, 周庆华, 吴迎, 等. 超微血管三维立体成像技术对乳腺肿块血流的评估价值[J]. 中国超声医学杂志, 2019, 35(7): 587-589. doi: 10.3969/j.issn.1002-0101.2019.07.005
[17]	蔡思曼, 王红燕, 张晓燕, 等. 智能三维超微血管成像评估乳腺病变血流的观察者间一致性分析[J]. 中华超声影像学杂志, 2020, 29(7): 613-617.
[18]	肖露, 褚雯, 王华. 超微血管成像技术对乳腺肿瘤血管形态分布特征及其诊断效能的初步分析[J]. 中华超声影像学杂志, 2018, 27(11): 973-976. https://cdmd.cnki.com.cn/Article/CDMD-10464-1019658790.htm
[19]	CHOI J S, HAN B K, KO E S, et al. Effect of a deep learning framework-based computer-aided diagnosis system on the diagnostic performance of radiologists in differentiating between malignant and benign masses on breast ultrasonography[J]. Korean J Radiol, 2019, 20(5): 749-758.

施引文献

资源附件(0)

图(1) / 表(2)

计量

文章访问数: 163
HTML全文浏览量: 37
PDF下载量: 19
被引次数: 0

人工智能联合超微血管成像技术在乳腺结节诊断中的价值

通讯作者: 笪应芬, E-mail: dayingfen@126.com

计量

出版历程

Value of artificial intelligence combined with super microvascular imaging technology in the diagnosis of breast nodules

计量

出版历程

目录

通讯作者:
笪应芬, E-mail: dayingfen@126.com