Abstract: We propose to develop, optimize and validate novel DW-MRI acquisition and modeling methods, which address non-Gaussian water diffusion and perfusion effects through diffusion kurtosis imaging and non- Gaussian intravoxel incoherent motion imaging and provide more specific measures of tissue structure and biology. Additionally, we will develop and implement advanced image processing tools to maximize the biologic information from the tumor/tissue provided by the imaging data. The essence of our timely proposal lies in it being the first multi-center, imaging trial to identify quantitative imaging biomarkers as early response to therapy indicators, which interrogate tumor biology in accordance with the central mission of the NCI Quantitative Imaging Network. It will address an urgent, unmet need in clinical trials for recurrent/metastatic (R/M) head and neck cancers. This UO1 proposal is in response to PAR-14-116 and the specific aims outlined in the proposal are as follows: Aim 1: To develop and standardize a multi b-value reduced field of view (rFOV) DW-MRI acquisition method and non-mono exponential modeling DW-MRI for oncology applications; Aim 2: To develop and implement optimal model methodology with advanced image segmentation and image feature analysis in patients with R/M malignancies in the HN region for oncology applications; and Aim 3: To establish the next generation DW-MRI biomarkers as early response to therapy indicators in experimental therapies using R/M HN squamous cell carcinoma (SCC) as a proof of principle model. We hypothesize that imaging metrics derived from newer methods can be used as quantitative imaging biomarkers for assessing early therapeutic efficacy in R/M HNSCC. The principles of identifying robust, reliable and quantitative imaging biomarkers derived from DW-MRI and image feature analysis remain similar and such imaging protocols, after appropriate adaptation, can have a wider clinical application, including their use in treating other solid tumors.