新辅助放化疗（nCRT）后达到病理学完全缓解的食管鳞癌患者会有显著的术后生存获益，如果能早期预测nCRT后的病理反应，则有助于可手术食管鳞癌患者的治疗决策，然而，目前还没有公认的准确预测食管鳞癌nCRT后病理反应分级的模型。基于前期研究发现PET和MRI功能影像能不同程度地预测食管鳞癌nCRT后病理反应，本项目拟以PET和MRI功能影像为基础，通过筛选和测试功能影像、纹理参数、临床因素的预测价值，建立更为准确有效的食管鳞癌nCRT后病理反应分级预测模型：①确定PET及MRI DWI图像所示代谢-细胞密度表型结合图像纹理参数对食管鳞癌nCRT后病理反应分级的预测价值；②基于裸鼠移植瘤模型确定PET及MRI DWI图像参数在食管鳞癌nCRT中各时间节点的变化与组织病理反应的关系；③将功能影像参数、图像纹理参数、临床因素纳入疗效预测模型，通过术后病理验证预测模型的准确性。

The patients with esophageal squamous cell carcinoma (ESCC) would benefit from neoadjuvant chemoradiotherapy (nCRT) when achieving pathologic complete response, significantly. The prediction of the pathological response grading in the early stage of treatment could contribute to make treatment decisions for the operable ESCC. However, there is still no concensus precise model for predicting the pathologic response grading for ESCC. Our previous study found that PET and MRI imaging could forecast the post-nCRT pathological response to different extend. This study aims to establish a more accurate and efficient model to predict the post-nCRT pathological response grading, by screening and testing the predictive value of PET and MRI functional imaging, texture parameters, and clinical features. Firstly, probing into the predictive value of metabolism and cell-density phenotype and texture parameters derived from PET and MRI diffusion weighted imaging (DWI) in the post-nCRT pathological response; secondly, confirming the correlation between the variation of PET/MRI DWI parameters in different phases of nCRT and the histopathologic response for ESCC, based on the nude-mouse transplanted tumor model; finally, integrating functional imaging, texture parameters, and clinical features into the predictive model, and verifying the accuracy of proposed model by postoperative pathology.