> Our Work > The Work We Fund

2021 Greg Brown in Honor of Plum and Jonathan W. Simons, MD-PCF Young Investigator Award

Automated Interpretation of PSMA and FDG PET Images to Improve Prognosis and Management in Advanced Prostate Cancer

Price Jackson, PhD
Peter MacCallum Cancer Centre

Mentors: Michael Hofman, MBBS, Kai Qin, PhD

Description:

  • PSMA-PET and FDG-PET are two molecular imaging technologies that have been used to aid selection of patients who are most likely to benefit from 177Lu-PSMA-targeted radionuclide therapy (LuPSMA), a new treatment that will likely soon be FDA approved for advanced prostate cancer patients.
  • An individual’s likelihood of response to LuPSMA treatment is related to PSMA levels and disease phenotypes seen on PSMA-PET, and concordance between PSMA & FDG PET imaging. However, visual assessment of these large datasets can be time-consuming and the human eye may fail to appreciate the complex changes that occur in response to specific therapies.
  • Dr. Price Jackson is developing an artificial intelligence (AI) machine-learning based approach to evaluate PSMA & FDG PET imaging data from patients treated with LuPSMA on a number of clinical trials conducted at the Peter MacCallum Cancer Centre.
  • Computational algorithms will be developed to automatically define total tumor volume, correlate imaging biomarkers and predict treatment outcome from a variety of therapeutic regimens.
  • Additionally, a similar technological approach will be developed to use SPECT imaging to track treatment responses and calculate the radiation doses delivered to tumor and organs following treatment with LuPSMA.
  • If successful, this project will result in an automated technology that can be used to identify, stage, and refer patients for the ideal cancer therapy based on their unique pattern of disease.

What this means to patients: PSMA-PET and FDG-PET can be used to select patients for treatment with LuPSMA. Dr. Jackson is developing an AI-based method to evaluate PSMA-PET and FDG-PET imaging data and identify patients most likely to benefit from LuPSMA. This automated computational method can be freely shared and will offer a consistent level of expertise across centers, enabling clinicians to optimally identify patients who should receive this promising emerging treatment.