能谱CT和人工智能在甲状腺癌诊断中的应用

Application of spectral CT and artificial intelligence in the diagnosis of thyroid cancer

甲状腺能谱CT扫描可减少图像伪影,降低噪声,提高对比噪声比及信噪比,同时通过多参数定量分析病灶特征和组成成分等,使得甲状腺癌的CT诊断水平有了显著的提高。近年来,医疗人工智能发展迅猛,在CT图像纹理分析、影像学定量特征提取和深度学习中充分突显高敏感检出、高维信息挖掘及高通量计算能力。在能谱CT基础上,辅助人工智能分析将有望为甲状腺癌术前提供更丰富、更精准的特征信息,不仅为早期准确诊断甲状腺癌及其转移性淋巴结评估提供更多参考价值,更为有效辅助临床诊断和实现个体化预后预测提供可行之路。因此,本文对甲状腺癌的能谱CT诊断技术和人工智能的研究应用进行综述。

Scanning of energy spectral computed tomography(CT)on thyroid can reduce image artifacts,lower noise,and enhance contrast-to-noise ratio and signal-to-noise ratio. At the same time,it can also quantitatively analyze lesion characteristics and composition through multiple parameters,resulting in a significant improvement in diagnosis of thyroid cancer on CT. Medical artificial intelligence has been developing rapidly in recent years,and it fully highlights high-sensitivity detection,high-dimensional information mining,and high-throughput computing capabilities in the aspects of CT image texture analysis,imaging quantitative feature extraction,and deep learning.Based on spectral CT,artificial intelligence-assisted analysis is expected to provide richer and more accurate characteristic information for preoperative preparation for thyroid cancer. It can not only provide more reference value for early diagnosis of thyroid tumors and assessment of lymph node metastasis,but also offer a feasible way to assist effectively in clinical diagnosis and to achieve individualized prognosis prediction. Therefore,this article summarizes the current research and application of spectral CT diagnosis technology and artificial intelligence of thyroid cancer.