A University of Sydney medical imaging expert is among a team of global researchers who have developed a new PET scanning technique that promises to dramatically decrease patients’ exposure to radiation.
PET scans – or positron emission tomography scans – are widely used to diagnose a variety of diseases, including many types of cancers, heart disease and Alzheimer’s disease.
A PET scan usually requires a small amount of liquid radioactive material to be injected into a patient’s bloodstream. This radioactive material accumulates in a person’s body where it gives off energy in the form of gamma rays. Gamma rays help doctors obtain high-quality images for clinical needs.
While previous research shows the danger of developing cancer from this radiation is low, the risks are higher for patients who receive multiple PET scans during their treatments, and for children.
Traditionally, clinicians have aimed to keep radiation doses as low as possible. The major drawback of dose reduction is that more visual distortion may be involved in the reconstructed PET images, resulting in poor image quality.
Now, a team of researchers from the University of Sydney, University of North Carolina at Chapel Hill, Sichuan University, the University of Wollongong, North Carolina State University, and Nanjing University of Aeronautics and Astronautics has proposed a novel method using unique 3D medical imaging techniques which reduce the radiation dose while maintaining the high quality of PET images.
The findings were recently published in NeuroImage.
One of corresponding authors Dr Luping Zhou, an expert in medical imaging from the University of Sydney’s School of Electrical and Information Engineering, and her colleagues tested their method on human brains, including subjects diagnosed with a mild cognitive impairment.
“Our experimental results show that our proposed method outperforms the benchmark methods and achieves much better performance than the state-of-the-art methods in both qualitative and quantitative measures – our method improves the clinical usability, processing speed and encourages less blurring and overall better image quality,” Dr Zhou said.
Different from other medical imaging techniques that consider an image’s appearance slice by slice, the researchers proposed method is 3D.
“Our technique uses unique machine learning algorithms – known as 3D conditional generative adversarial networks (or 3D c-GANs) – to estimate the high-quality full-dose PET images from low-dose ones,” Dr Zhou said.