Objective: To investigate the value of ultrasonic images texture analysis in predicting the metastasis of cervical lymph node for thyroid carcinoma. Methods: A total of 106 thyroid carcimoma patients were enrolled in this study. All patients underwent cervical lymph node dissection and confirmed by pathology. Then the patients were divided into two groups, including metastasis group and non-metastasis group. The ultrasonic images of the thyroid carcinama were imported into Mazda 4.6 software and reqin of interest (ROIs) were manually drawed. The optimum texture parameters were selected through Fisher coefficient, probability of classification error and average correction coefficient (POE+ACC), Mutual information (MI) and the combination of three methods (Fisher+POE+ACC+MI) respectively.The model of artificial neural network was builded. The difference of texture features between metastasis group and non-metastasis group was compared. And misdiagnostic rates were analyzed between texture analysis and ultrasonic doctors. Results: There was no significant difference in misdiagnosti rates between Fisher coefficient, POE+ACC, MI, Fisher+Poe+ACC+MI and ultrasound doctors (χ2 values were 2.652, 2.141, 0.897, 0.175, P>0.05). The misdiagnosis rate of ultrasound doctors in the metastasis group was significantly higher than that in the non metastasis group(χ2=14.265, P<0.05), but in the metastasis group the misdiagnostic rates of MI and three methods combination were significantly lower than that of ultrasound doctors (χ2 values were 5.032, 8.705, P<0.05). Conclusion: Ultrasonic images texture analysis has a good predictive value for cervical lymph node metastasis of thyroid carcinoma. |
[1] NOCINI R,MOLTENI G,MATTIUZZI C,et al.Updates on larynx cancer epidemiology[J].Chinese J Cancer Res,2020,32(1):18-25.
[2] 徐丹凤,王志利,刘洋,等.超声弹性成像联合TSH检测对甲状腺结节良恶性鉴别诊断的价值分析[J].东南大学学报(医学版),2018,37(1):50-54.
[3] WEI X,LI D,GAO M,et al.Prediction of cervical lymph node metastases in papillary thyroid microcarcinoma by sonographic features of the primary site[J].Cancer Biol Med,2019,16(3):587-594.
[4] 袁文利,王玉梅,王勇,等.甲状腺癌的超声影像学诊断与病理学对照研究[J/OL].中华普外科手术学杂志(电子版),2018,12(3):63-66.
[5] 刘军,李新苹,苏磊,等.206例甲状腺乳头状癌患者颈侧区淋巴结转移临床分析[J].东南大学学报(医学版),2020,39(1):70-73.
[6] 中华人民共和国国家卫生健康委员会.甲状腺癌诊疗规范(2018年版)[J/OL].中华普通外科学文献(电子版),2019,13(1):1-15.
[7] 郑璇,陈松旺.普通超声与实时弹性超声在诊断甲状腺癌转移性颈部淋巴结中的应用[J].中国医疗设备,2015,23(6):62-66.
[8] KASSNER A,THORNHILL R E.Texture analysis:a review of neurologic MR imaging applications[J].AJNR Am J Neuroradiol,2010,31(5):809-816.
[9] 徐琰,胡保全.浅谈人工智能在乳腺癌领域的应用进展[J/OL].中华乳腺病杂志(电子版),2017,19(5):7-11.
[10] HOUSSAMI N,LEE C I,BUIST D S M,et al.Artificial intelligence for breast cancer screening:Opportunity or hype[J].Breast,2018,36(12):31-33.
[11] 熊伟,龚勋,罗俊,等.基于局部纹理特征的超声甲状腺结节良恶性识别[J].数据采集与处理,2015,30(1):186-191.
[12] 种美玲,时白雪,张禧,et al.超声联合纹理分析对乳腺结节良恶性的诊断价值[J/OL].中华医学超声杂志(电子版),2019,16(8):581-585.
[13] 关欣,吴鸿莉.彩色多普勒超声图像纹理分析与超声造影对肝占位性病变鉴别诊断价值对比探究[J].创伤与急危重病医学,2019,7(6):359-361,366.
[14] 谭红娜,武明辉,顾建钦,等.乳腺X线图像纹理特征预测乳腺癌腋窝淋巴结转移[J].中国医学影像技术,2017,33(12):1774-1778.
[15] 楼飞.超声诊断甲状腺癌颈部淋巴结转移区域及特点的价值[J].浙江临床医学,2019,21(8):1133-1134.
[16] 张迅,曾炜,陈雅玲,等.超声评价甲状腺癌颈部淋巴结转移的价值[J].中华超声影像学杂志,2011,20(9):825-826. |