Described herein are particle-driven radiogenomics systems and methods that can be used to identify imaging features for prediction of intratumoral and interstitial nanoparticle distributions in cancers (e.g., in low grade and/or high-grade brain cancers (e.g., gliomas, e.g., primary gliomas)). In certain embodiments, the systems and methods described herein extract and combine quantitative multi-dimensional data generated from structural, functional, and/or metabolic imaging. In certain embodiments, the combined multidimensional data is linked to intratumoral and interstitial nanoparticle distributions. For example, this linked data can be used to determine quantitative functional-metabolic multimodality particle-based imaging features and to predict treatment efficacy. These techniques provide an improved quantitative ability to measure treatment response and determine tumor progressions compared to traditional size-based imaging methods.
