AI辅助多模态超声联合血清促甲状腺激素对甲状腺结节定性的诊断效能

托静美; 司晓娟; 宋和琴

doi:10.16766/j.cnki.issn.1674-4152.003723

AI辅助多模态超声联合血清促甲状腺激素对甲状腺结节定性的诊断效能

doi: 10.16766/j.cnki.issn.1674-4152.003723

1.
张家口市第一医院超声医学科，河北张家口 075000
2.
张家口市第一医院检验医学科

基金项目:

河北省卫生健康委员会医学科学研究课题 20221902

详细信息

通讯作者:
托静美，E-mail: 17832331555@163.com

中图分类号: R445.1 R581.9
计量
- 文章访问数: 13
- HTML全文浏览量: 12
- PDF下载量: 2
- 被引次数: 0
出版历程
- 收稿日期: 2024-04-18
- 网络出版日期: 2024-12-28

Qualitative efficacy of AI-assisted multimodal ultrasound combined with serum thyroid stimulating hormone in guiding thyroid nodules

1.
Department of Ultrasound Medicine, Zhangjiakou First Hospital, Zhangjiakou, Hebai 075041, China

摘要

摘要: 目的探究甲状腺结节定性诊断中人工智能(AI)辅助多模态超声联合血清促甲状腺激素(TSH)的指导效能，以期为甲状腺结节临床诊治提供参考。方法选取2021年1月—2022年6月张家口市第一医院就诊的甲状腺结节性疾病患者100例，采用AI常规超声、超声造影、实时剪切波弹性成像(SWE)检查，以病理结果为金标准，将患者分为良性组(23例)与恶性组(77例)。检测2组血清TSH水平，采用ROC曲线分析TSH对甲状腺结节定性的预测价值，并评估AI辅助多模态超声联合TSH对甲状腺结节定性的诊断效能。结果检查发现113个结节，其中26个良性，87个恶性。AI常规超声+超声造影+SWE联合诊断的灵敏度为89.66%(78/87)，比较单一及两两联合诊断高(P＜0.05)。良性组血清TSH水平比恶性组低(P＜0.05)。ROC曲线显示，血清TSH诊断的AUC、灵敏度、特异度、截断值分别为0.800、74.71%、84.62%、4.44 mIU/L。AI辅助多模态超声联合血清TSH诊断的准确度、灵敏度、阴性预测值分别为93.81%(106/113)、97.70%(85/87)、91.30%(21/23)，比较单一诊断高(P＜0.05)。结论在甲状腺结节定性诊断中AI辅助多模态超声联合血清TSH的指导效能确切，应用价值较高。
- 甲状腺结节 /
- 超声 /
- 人工智能 /
- 促甲状腺激素
Abstract: Objective To explore the guiding efficacy of artificial intelligence (AI) assisted multimodal ultrasound combined with serum thyroid stimulating hormone (TSH) in qualitative diagnosis of thyroid nodules, in order to provide reference for clinical diagnosis and treatment of thyroid nodules. Methods A total of 100 patients with thyroid nodular disease treated in Zhangjiakou First Hospital from January 2021 to June 2022 were selected and examined by AI routine ultrasound, contra-ultrasound, and real-time shear wave elastography (SWE). The patients were divided into a benign group (23 cases) and a malignant group (77 cases) with pathological results as the gold standard. Serum TSH levels were detected in the two groups. The receiver operating characteristic curve (ROC) was used to analyze the predictive value of TSH in the qualitative diagnosis of thyroid nodules, and the diagnostic efficacy of AI-assisted multimodal ultrasound combined with TSH in the qualitative diagnosis of thyroid nodules was evaluated. Results A total of 113 nodules were found by pathological findings, including 26 benign nodules and 87 malignant nodules. The sensitivity of AI conventional ultrasound + CEUS +SWE combined diagnosis was 89.66% (78/87), which was higher than that of single and pairwise combined diagnosis (P < 0.05). The serum TSH level in the benign group was lower than that in the malignant group (P < 0.05). The ROC curve showed that the AUC (95% CI), sensitivity, specificity, and cut-off values of serum TSH diagnosis were 0.800, 74.71%, 84.62%, and 4.44 mIU/L, respectively. The accuracy, sensitivity, and negative predictive value of AI-assisted multimodal ultrasound combined with serum TSH diagnosis were 93.81% (106/113), 97.70% (85/87), and 91.30% (21/23), respectively, which were higher than that of single diagnosis (P < 0.05). Conclusion AI-assisted multimodal ultrasound combined with serum TSH is effective and valuable in the qualitative diagnosis of thyroid nodules.
- Thyroid nodules /
- Ultrasound /
- Artificial intelligence /
- Thyroid stimulating hormone

HTML全文

图 1 血清TSH对良恶性结节诊断的ROC曲线

Figure 1. ROC curve of serum TSH in diagnosis of benign and malignant nodules

下载: 全尺寸图片幻灯片

图 2 典型病例影像学及病理图片

注：A、B、C为二维超声图片，D、E为AI常规超声图片，F、G为SWE技术图片，H、I为病理图片。

Figure 2. Imaging and pathological images of typical cases

下载: 全尺寸图片幻灯片

表 1 不同方法诊断结果与病理结果比较(个)

Table 1. Comparison of diagnostic results and pathological results by different methods (cases)

方法	病理结果			Kappa值	P值
方法	恶性	良性	合计	Kappa值	P值
AI常规超声				0.463	＜0.001
恶性	64	4	68
良性	23	22	45
合计	87	26	113
超声造影				0.434	＜0.001
恶性	60	3	63
良性	27	23	50
合计	87	26	113
SWE				0.474	＜0.001
恶性	63	3	66
良性	24	23	47
合计	87	26	113
AI常规超声+超声造影				0.492	＜0.001
恶性	66	4	70
良性	21	22	43
合计	87	26	113
AI常规超声+SWE				0.496	＜0.001
恶性	68	5	73
良性	19	21	40
合计	87	26	113
AI常规超声+超声造影+SWE				0.611	＜0.001
恶性	78	7	85
良性	9	19	28
合计	87	26	113

下载: 导出CSV

表 2 不同方法对甲状腺结节定性诊断的诊断效能比较(%)

Table 2. Comparison of diagnostic effectiveness of different methods for qualitative diagnosis of thyroid nodules (%)

方法	准确度	灵敏度	特异度	阳性预测值	阴性预测值
AI常规超声	76.11(86/113)	73.56(64/87)^a	84.62(22/26)	94.12(64/68)	48.89(22/45)
超声造影	73.45(83/113)^a	68.97(60/87)^a	88.46(23/26)	95.24(60/63)	46.00(23/50)
SWE	76.11(86/113)	72.41(63/87)^a	88.46(23/26)	95.45(63/66)	48.94(23/47)
AI常规超声+超声造影	77.88(88/113)	75.86(66/87)^a	84.62(22/26)	94.29(66/70)	51.16(22/43)
AI常规超声+SWE	78.76(89/113)	78.16(68/87)^a	80.76(21/26)	93.15(68/73)	52.50(21/40)
AI常规超声+超声造影+SWE	85.84(97/113)	89.66(78/87)	73.08(19/26)	91.76(78/85)	67.86(19/28)
χ²值	5.927	12.476	3.138	1.244	3.867
P值	0.313	0.029	0.679	0.941	0.569
注：与AI常规超声+超声造影+SWE比较，^aP＜0.05。

下载: 导出CSV

表 3 AI辅助多模态超声、血清TSH单一及联合诊断结果与病理结果比较(个)

Table 3. Comparison of AI-assisted multimodal ultrasound, serum TSH single and combined diagnosis and pathological results(cases)

方法	病理结果			Kappa值	P值
方法	恶性	良性	合计	Kappa值	P值
TSH				0.477	＜0.001
恶性	65	4	69
良性	22	22	44
合计	87	26	113
AI常规超声+超声造影+SWE				0.611	＜0.001
恶性	78	7	85
良性	9	19	28
合计	87	26	113
AI常规超声+超声造影+SWE+TSH				0.818	＜0.001
恶性	85	5	90
良性	2	21	23
合计	87	26	113

下载: 导出CSV

表 4 AI辅助多模态超声、血清TSH单一及联合对甲状腺结节定性诊断的诊断效能比较(%)

Table 4. Comparison of diagnostic efficacy of AI-assisted multimodal ultrasound, serum TSH single and combined in qualitative diagnosis of thyroid nodules (%)

方法	准确度	灵敏度	特异度	阳性预测值	阴性预测值
TSH	76.99(87/113)^a	74.71(65/87)^ab	84.62(22/26)	94.20(65/69)	50.00(22/44)^a
AI常规超声+超声造影+SWE	85.84(97/113)^a	89.66(78/87)^a	73.08(19/26)	91.76(78/85)	67.86(19/28)^a
AI常规超声+超声造影+SWE+TSH	93.81(106/113)	97.70(85/87)	80.77(21/26)	94.44(85/90)	91.30(21/23)
χ²值	12.930	21.438	1.101	0.603	11.485
P值	0.002	＜0.001	0.577	0.740	0.003
注：与AI常规超声+超声造影+SWE+TSH比较，^aP＜0.05；与AI常规超声+超声造影+SWE比较，^bP＜0.05。

下载: 导出CSV

参考文献(22)

[1]	单忠艳. 长期全民食盐加碘的有效性和安全性: 中国大陆31省流行病学证据[J]. 国际内分泌代谢杂志, 2020, 40(5): 314. SHAN Z Y. Effectiveness and safety of long-term national salt iodization: Epidemiological evidence in 31 provinces in China[J]. International Journal of Endocrinology and Metabolism, 2020, 40(5): 314.
[2]	李艳辉, 赵晨希, 陈明, 等. 超声检查在甲状腺大结节成年患者中的应用价值[J]. 中华全科医学, 2022, 20(10): 1742-1745, 1808. doi: 10.16766/j.cnki.issn.1674-4152.002693 LI Y H, ZHAO C X, CHEN M, et al. Application value of ultrasound in adult patients with thyroid nodules≥3 cm[J]. Chinese Journal of General Practice, 2022, 20(10): 1742-1745, 1808. doi: 10.16766/j.cnki.issn.1674-4152.002693
[3]	麦武平, 周美君, 胡叶, 等. AI超声辅助诊断系统在低年资医师诊断甲状腺结节良恶性中的应用[J]. 中国超声医学杂志, 2021, 37(3): 248-251. MAI W P, ZHOU M J, HU Y, et al. The Application of AI Ultrasound-assisted Diagnosis System in Diagnosing Benign and Malignant Thyroid Nodules for Junior Sonographers[J]. Chinese Journal of Ultrasound in Medicine, 2021, 37(3): 248-251.
[4]	PENG S, LIU Y, LV W, et al. Deep learning-based artificial intelligence model to assist thyroid nodule diagnosis and management: a multicentre diagnostic study[J]. Lancet Digit Health, 2021, 3(4): e250-e259. doi: 10.1016/S2589-7500(21)00041-8
[5]	戴红, 李慎义, 龙湘党, 等. 超声造影定量参数联合血清TSH, Tg, 外周血NLR在甲状腺结节良恶性诊断中的应用价值[J]. 现代生物医学进展, 2023, 23(11): 2094-2097. DAI H, LI S Y, LONG X D, et al. Application Value of Contrast-Enhanced Ultrasound Quantitative Parameters Combined with Serum TSH, Tg, Peripheral Blood NLR in the Diagnosis of Benign and Malignant Thyroid Nodules[J]. Advances Modern Biomedi, 2023, 23(11): 2094-2097.
[6]	陈晨, 孙鹏飞, 郭君, 等. 人工智能在甲状腺结节良恶性中的诊断价值[J]. 中国超声医学杂志, 2020, 36(7): 585-588. CHEN C, SUN P F, GUO J, et al. Diagnostic Value of Artificial Intelligence in Benign and Malignant Thyroid Nodules[J]. Chinese Journal of Ultrasound in Medicine, 2020, 36(7): 585-588.
[7]	周静, 徐秀梅, 翟虹. 超声造影对不同大小TI-RADS 4类甲状腺结节的定性诊断价值[J]. 新疆医科大学学报, 2022, 45(10): 1146-1150. ZHOU J, XU X M, ZHAI H. Qualitative diagnostic value of contrast-enhanced ultrasound in TI-RADS 4 thyroid nodules of different sizes[J]. Journal of Xinjiang Medical University, 2022, 45(10): 1146-1150.
[8]	梁萍, 姜玉新. 超声E成像临床应用指南[M]. 北京: 人民卫生出版社, 2018. LIANG P, JIANG Y X. Guidelines for clinical application of ultrasound E imaging[M]. Beijing: People ' s Medical Publishing House, 2018.
[9]	卢景新, 陈俊君, 何志忠, 等. 超声造影对C-TIRADS中甲状腺4~5类结节良恶性的鉴别价值[J]. 贵州医科大学学报, 2022, 47(6): 704-709. LU J X, CHEN J J, HE Z Z, et al. Diagnostic value of contrast-enhanced ultrasound in differentiating benign and malignant thyroid nodules in C-TIRADS 4-5[J]. Journal of Guizhou Medical University, 2022, 47(6): 704-709.
[10]	CHEN F, HAN H, WAN P, et al. Joint segmentation and differential diagnosis of thyroid nodule in contrast-enhanced ultrasound images[J]. IEEE Trans Biomed Eng, 2023, 70(9): 2722-2732. doi: 10.1109/TBME.2023.3262842
[11]	陈佳慧, 张宇晴, 朱天彤, 等. 超声造影在不同大小甲状腺结节鉴别诊断中的研究[J]. 中国临床医学影像杂志, 2022, 33(12): 842-853. CHEN J H, ZHANG Y Q, ZHU T T, et al. Contrast-enhanced ultrasound in differential diagnosis of thyroid nodules with different sizes[J]. Journal of China Clinic Medical Imaging, 2022, 33(12): 842-853.
[12]	TUAN P A, DUC N M, AN M, et al. The role of shear wave elastography in the discrimination between malignant and benign thyroid nodules[J]. Acta Inform Med, 2020, 28(4): 248-253. doi: 10.5455/aim.2020.28.248-253
[13]	陈智, 杨帆. SWE定量参数在甲状腺结节定性诊断中的价值及与肿瘤恶性程度的相关性[J]. 南昌大学学报: 医学版, 2022, 62(5): 58-62. CHEN Z, YANG F. Value of SWE Quantitative Parameters in Qualitative Diagnosis of Thyroid Nodules and Their Relationship to Tumor Malignancy[J]. Journal of Nanchang University: Medical Edition, 2022, 62(5): 58-62.
[14]	崔楠, 陈雨凡, 柏通, 等. 甲状腺结节AI超声辅助诊断系统在住院医师规范化培训教学中的应用[J]. 中国医学物理学杂志, 2023, 40(10): 1233-1236. CUI N, CHEN Y F, BAI T, et al. Application of ACR TI-RADS-based computer-aided diagnosis system for thyroid nodules in the teaching practice of standardized training for resident doctors[J]. Chinese Journal of Medical Physics, 2023, 40(10): 1233-1236.
[15]	WILDMAN-TOBRINER B, TAGHI-ZADEH E, MAZUROWSKI M A. Artificial intelligence (AI) tools for thyroid nodules on ultrasound, from the AJR special series on AI applications[J]. AJR Am J Roentgenol, 2022, 219(4): 1-8.
[16]	陆海永, 李朝喜, 温德惠, 等. 多模态超声在中、高度可疑恶性甲状腺结节鉴别中的诊断价值[J]. 天津医药, 2021, 49(1): 69-73. LU H Y, LI Z X, WEN D H, et al. Diagnostic values of multimodal ultrasound imaging in the differentiation of moderate and highly suspected malignant thyroid nodules[J]. Tianjin Medical Journal, 2021, 49(1): 69-73.
[17]	龚忠静, 杨慧娴, 杨青, 等. AI超声智能辅助诊断系统对甲状腺结节诊断价值的初步探讨[J]. 中国临床医学影像杂志, 2021, 32(11): 781-784, 788. GONG Z J, YANG H X, YANG Q, et al. Preliminary study of diagnostic value of AI intelligent assistant diagnostic system for thyroid nodules[J]. Journal of China Clinic Medical Imaging, 2021, 32(11): 781-784, 788.
[18]	曹毅, 李斌, 陈柯余, 等. 女性甲状腺结节与生活及月经生殖因素的相关性分析[J]. 中国慢性病预防与控制, 2020, 28(9): 692-695. CAO Y, LI B, CHEN K Y, et al. Correlation analysis between thyroid nodule and life, menstrual and reproductive factors in female[J]. Chinese Journal of Prevention and Control of Chronic Diseases, 2020, 28(9): 692-695.
[19]	李炳荣, 王海林, 周利民. 双源CT联合血清TSH在判断甲状腺结节性质中的应用价值[J]. 新医学, 2019, 50(6): 437-441. LI B R, WANG H L, ZHOU L M. Application value of dual-source CT combined with serum thyroid-stimulating hormone in diagnosis of thyroid nodules[J]. New Medicine, 2019, 50(6): 437-441.
[20]	王欣, 赵玉珠. 促甲状腺激素与甲状腺癌发病及病理特征的关系研究[J]. 中国实用内科杂志, 2021, 41(5): 424-427. WANG X, ZHAO Y Z. Relationship between TSH level and the incidence and pathological characteristics of thyroid cancer[J]. Chinese Journal of Practical Internal Medicine, 2021, 41(5): 424-427.
[21]	丁智慧. 血清甲状腺球蛋白抗体甲状腺过氧化物酶抗体促甲状腺激素检测在良恶性甲状腺结节鉴别诊断中的研究[J]. 山西医药杂志, 2023, 52(3): 225-228. DING Z H. Detection of serum thyroglobulin antibody thyroperoxidase antibody thyrotropin in differential diagnosis of benign and malignant thyroid nodules[J]. Shanxi Medical Journal, 2023, 52(3): 225-228.
[22]	宋和琴, 托静美, 司晓娟. 多模态超声定量分析联合血清TSH, TRAb水平对甲状腺结节良恶性的诊断价值[J]. 东南大学学报: 医学版, 2023, 42(5): 730-735. SONG H Q, TUO J M, SI X J. Diagnostic value of multimodal ultrasound quantitative analysis combined with serum TSH and TRAb levels for benign and malignant thyroid nodules[J]. Journal of Southeast University(Medical Science Edition), 2023, 42(5): 730-735.