基于CT影像及组学特征评估肺磨玻璃结节病理分级的研究进展

Research Progress on Pathological Grading of Pulmonary Ground-Glass Nodules Based on CT Imaging and Omics Features

摘要: 肺磨玻璃结节(GGN)是一种肺部常见的非特异性征象，良恶性GGN的治疗方法及预后不同，因此准确鉴别GGN具有重要的临床意义。影像组学可以从影像图像中高通量地提取影像特征，并从中推断出可能包含预后信息的基因蛋白表型或特征。CT影像组学可以鉴别GGN的良恶性，预测恶性GGN的病理分型并评估其侵袭性，为肺GGN个体化诊疗方案的选择提供有力的依据。为此本文将综述传统CT形态学特征、影像组学定量参数在GGN病理分级中的应用，多参数融合预测模型的构建现状，分析模型泛化能力不足、临床转化困难等当前研究挑战，并展望标准化扫描方案、多组学联合等未来发展趋势，旨在为临床实现GGN的个体化精准管理提供理论参考。

Abstract: Pulmonary ground-glass nodules (GGNs) are a common nonspecific pulmonary finding. The treatment and prognosis of benign and malignant GGNs differ, making accurate differentiation clinically significant. Imaging omics can extract high-throughput imaging features from images and infer gene/protein phenotypes or characteristics potentially containing prognostic information. CT imaging omics can distinguish benign from malignant GGNs, predict pathological subtypes of malignant GGNs, and evaluate their invasiveness, providing robust evidence for personalized treatment strategies. This review summarizes the application of traditional CT morphological features and imaging omics quantitative parameters in GGN pathological grading, the current status of multi-parameter fusion prediction models, and analyzes existing research challenges such as limited model generalization and clinical translation difficulties. Future trends include standardized scanning protocols and multi-omics integration, aiming to provide theoretical references for clinical precision management of GGNs.