An information-theoretic approach to HP signal acquisition in brain

Project Summary Dissolution dynamic nuclear polarization can be used to enhance the MRI signal from [1-13C]-pyruvate by 4-5 orders of magnitude compared to thermal equilibrium, enabling imaging of hyperpolarized (HP) pyruvate and its metabolites with unprecedented spatiotemporal resolution.

The conversion of HP pyruvate into lactate is partic- ularly interesting in oncology because it provides direct insight into the Warburg effect, the metabolic phenotype of many cancers that produce lactate even under normoxic conditions.

Since the safety and feasibility of HP MRI was demonstrated in patients with prostate cancer, ten sites worldwide are now using this technology to assess changes in metabolism in human subjects, including normal brain metabolism and patients with brain cancer. This technology can provide unique new information to guide tissue biopsy.

However, formulating optimal HP acquisition strategies is exceptionally challenging. This project will develop an information-theoretic approach toward optimizing the data acquisition strategy.

We will develop optimal strategies for data sampling and anal- ysis of dynamic multispectral hyperpolarized imaging data.

The key idea is to identify measurements that are information-rich with respect to our biophysical model of the imaging signal. This signal recovery will be validated in clinical trials using HP imaging correlations with pathology.

Project outcomes are expected to improve the spatial accuracy in locating both low-grade and high-grade malignant glioma in the brain.