论文部分内容阅读
目的:运用BP神经网络技术建立甲状腺癌的无创诊断模型,评估该模型的预测诊断价值。方法:回顾性分析经术后病理证实为甲状腺癌39例与良性病变11例,提取出以上50例病例中手术前经过B超检查与甲状腺癌相关的8项图形特征,并进行评分量化,利用BP神经网络对50例病例进行学习和检验,建立甲状腺癌无创诊断模型。用该无创诊断模型对疑为甲状腺癌20例患者进行术前预测并与术后病理进行比较。结果:本文所建立的基于BP神经网络技术的无创诊断模型在甲状腺癌及甲状腺良性病变的预测诊断中达到了100%的准确率。结论:基于BP神经网络技术的无创诊断模型,在甲状腺癌及良性病变的预测诊断中具有较高的应用价值,这无疑对辅助B超诊断甲状腺良恶性病变提供了新的技术支撑和研究思路。
OBJECTIVE: To establish a noninvasive diagnosis model of thyroid cancer using BP neural network technology and evaluate its predictive value. Methods: Thirty-nine cases of thyroid carcinoma and 11 cases of benign lesions confirmed by postoperative pathology were retrospectively analyzed. Eight of the above 50 surgically extracted thyroid cancer-related features of B-ultrasound were extracted and quantified. BP neural network in 50 cases of learning and testing, the establishment of noninvasive diagnosis of thyroid cancer model. Using this noninvasive diagnostic model, 20 patients with suspected thyroid cancer were preoperatively predicted and compared with postoperative pathology. Results: The noninvasive diagnostic model based on BP neural network established in this paper achieves a 100% accuracy in the predictive diagnosis of thyroid cancer and benign thyroid lesions. Conclusion: The noninvasive diagnostic model based on BP neural network technology has high value in predicting thyroid cancer and benign lesions. It undoubtedly provides new technical support and research ideas for the diagnosis of benign and malignant thyroid lesions by auxiliary B ultrasound.