建立并验证基于 delta 放射组学的模型，用于预测诱导化疗 (IC) 后局部晚期鼻咽癌 (LA-NPC) 患者的无进展生存期 (PFS)。总共 250 名 LA-NPC 患者（训练队列） ：n = 145；验证队列：n = 105）已入组。从 IC 前后进行的 MRI 扫描中提取放射组学特征，并计算这些特征的变化。特征选择后，使用 LASSO-Cox 回归分析构建了 delta 放射组学特征。开发并评估了包含独立临床指标和 delta 放射组学特征的预后列线图，以进行校准和区分。使用 Kaplan-Meier 方法评估列线图的风险分层。由 12 个特征组成的 delta 放射组学特征与预后独立相关。列线图整合了 delta 放射组学特征和临床因素，表现出出色的校准和辨别能力。该模型在训练队列中的 Harrell 一致性指数（C 指数）为 0.848，在验证队列中为 0.820。风险分层确定了两组 PFS 率显着不同的组。 IC后接受同步放化疗（CCRT）或放疗加辅助化疗（RT AC）的高危患者的三年PFS显着高于仅接受RT的患者，达到统计学意义。相比之下，对于低风险患者，接受 CCRT 或 RT AC 的患者 IC 后的三年 PFS 略高于仅接受 RT 的患者；然而，这种差异并未达到统计学显着性。我们基于 delta MRI 的放射组学模型可用于预测 PFS，并可指导 LA-NPC IC 后的后续治疗决策。版权所有 © 2024 Elsevier Ltd。保留所有权利。

To establish and validate a delta-radiomics-based model for predicting progression-free survival (PFS) in patients with locoregionally advanced nasopharyngeal carcinoma (LA-NPC) following induction chemotherapy (IC).A total of 250 LA-NPC patients (training cohort: n = 145; validation cohort: n = 105) were enrolled. Radiomic features were extracted from MRI scans taken before and after IC, and changes in these features were calculated. Following feature selection, a delta-radiomics signature was constructed using LASSO-Cox regression analysis. A prognostic nomogram incorporating independent clinical indicators and the delta-radiomics signature was developed and assessed for calibration and discrimination. Risk stratification by the nomogram was evaluated using Kaplan-Meier methods.The delta-radiomics signature, consisting of 12 features, was independently associated with prognosis. The nomogram, integrating the delta-radiomics signature and clinical factors demonstrated excellent calibration and discrimination. The model achieved a Harrell's concordance index (C-index) of 0.848 in the training cohort and 0.820 in the validation cohort. Risk stratification identified two groups with significantly different PFS rates. The three-year PFS for high-risk patients who received concurrent chemoradiotherapy (CCRT) or radiotherapy plus adjuvant chemotherapy (RT+AC) after IC was significantly higher than for those who received RT alone, reaching statistical significance. In contrast, for low-risk patients, the three-year PFS after IC was slightly higher for those who received CCRT or RT+AC compared to those who received RT alone; however, this difference did not reach statistical significance.Our delta MRI-based radiomics model could be useful for predicting PFS and may guide subsequent treatment decisions after IC in LA-NPC.Copyright © 2024 Elsevier Ltd. All rights reserved.