Automatic, Rapid, High-Resolution Metabolic Imaging for Improved Management of Patients with Glioma

Project Title: Automatic, Rapid, High-Resolution Metabolic Imaging for Improved Management of Patients with Glioma

Funded by Department of Defense, USAMRAA (2023-2027) 

Principal Investigator: Yan Li

Co-Investigator: Esin Ozturk Isik

Researchers: Abdullah Baş

Project Summary: Gliomas are the most common malignant primary brain tumors in adults. Their highly infiltrative nature makes it difficult to define and treat the full extent of these tumors using conventional imaging because they are not specific to tumor cells. Both standard of care and experimental therapies can exacerbate this effect, often resulting in ambiguous findings on conventional MR images that hinder accurate response assessments. Metabolic alterations often precede anatomic and microstructural changes, potentially enabling earlier intervention for treatment modification. Our previous studies have shown that metrics from MR spectroscopic imaging (MRSI) were able to differentiate tumor cells from normal brain, classify subtypes, and even predict survival. These metabolic parameters can reduce ambiguities in interpreting changes observed on conventional anatomic images and hence aid in the definition of response to therapy. Despite these benefits, MRSI has yet to be adopted into routine clinical practice due to long scan times, complicated acquisition schemes, lower spatial resolution, and a lack of automated pipelines to generate images of brain metabolism rapidly on scanners.

Keywords: Brain imaging, MRSI, Deep Learning

Objectives: 

  • To develop a fully automated clinical workflow for fast, whole-brain, millimeter resolution MRSI that will provide spatially characterized brain metabolism when deployed in the clinic.

Specific Aims:

  1.  To develop strategies for acquiring accelerated, high spatial resolution metabolic
  2. To develop a fully automated post-processing workflow for spectral reconstruction,
  3. To evaluate the proposed tools on patient management.

Collaboration/Translational Aspects:

This proposal will be directed by Drs. Li (an expert in MRS acquisition, processing, and application) and Lupo (an expert in applying AI to MRI methods & gliomas) at UCSF, Dr. Ozhinksy (an expert in MRSI prescription & AI) at the VA/UCSF, and Dr. Ozturk-Isik (an expert in MRSI reconstruction) at Bogazici University.