NPRE researcher teams with Carle Hospital to improve quantitative accuracy in molecular imaging

9/11/2017 Susan Mumm, Editor

Written by Susan Mumm, Editor

NPRE researcher teams with Carle Hospital to improve quantitative accuracy in molecular imaging

Assistant Prof. Shiva Abbaszadeh has teamed with the Carle Foundation Hospital to use active cancer patient scans to improve the accuracy of molecular imaging technology for disease diagnostics and treatment.

Abbaszadeh, of Nuclear, Plasma, and Radiological Engineering at Illinois, is working with radiologist Brett Yockey, a Carle affiliate. Their project, “Improving quantitative molecular imaging accuracy in clinical practice and assessing response to therapy,” has been awarded a one-year, $50,000 grant from the Carle Illinois Collaborative Research Seed Program of the Interdisciplinary Health Sciences Institute at Illinois.

Health care professionals use positron emission tomography (PET) scans to determine cancer patients’ Standard Uptake Values, metrics that doctors use to evaluate and diagnose the disease for treatment.

“The accurate quantification of the (SUV) value depends on so many factors, including the imaging modality used in the hospital, the detector technology used, and image reconstruction algorithms” Abbaszadeh said. “This value is really important but the accuracy is related to these parameters. I’m working to reduce uncertainty and improve quantitative accuracy.”

Actual cancer patient scans allow Abbaszadeh to investigate and model the effect of patient individuality and patient preparation, system hardware, and software. This is particularly important, she said, if a patient comes back for follow-up scans. Her models take into account variations and ways in which they have been adjusted over time.

“If we bring in a new system (for scanning), I would change the parameters in my model to see if the same patient data would show a difference between the current model and the previous one,” she said. “We want to evaluate whether the changes (in imaging) are occurring because of the technology or because of a change in the patient.”

The models may also help determine whether image variations reflect biological influences other than cancer progression. For example, changes in the scans could reflect chemotherapy interference with kidneys or liver function rather than cancer advancement.

“This is a practical problem that is really helping the healthcare professional," Abbaszadeh said.

 


Share this story

This story was published September 11, 2017.