Lower radiation imaging for monitoring melanoma
An AI-based solution enabling safer, lower radiation imaging without compromising diagnostic accuracy when monitoring melanoma is being developed, thanks to Monash Institute of Medical Engineering (MIME) project grant support.
Melanoma is the most dangerous form of skin cancer and people with advanced disease require frequent PET scans to track disease progression and treatment response—especially in people undergoing immunotherapy. While full-dose PET scans provide the high-quality imaging needed for accurate diagnosis, a low dose solution would reduce radiation exposure, improve cost effectiveness and environmental impacts.
Lead researchers, Professor Victoria Mar from Alfred Health and Associate Professor Zhaolin Chen from Monash University, Faculty of Information Technology have applied advanced machine learning algorithms to enhance the clarity of low-dose scans, allowing clinicians to monitor patients effectively while reducing radiation exposure.
“There are significant benefits to being able to provide high quality scans using a fraction of the radiation dose for patients requiring regular PET imaging, such as those with melanoma undergoing immunotherapy,” said Professor Victoria Mar and Associate Professor Zhaolin Chen.
“Our research addresses this critical gap by enabling safe, lower-radiation imaging without compromising diagnostic accuracy.
“This innovation aims to improve patient safety, enable more dynamic imaging and clinical decision-making whilst also reducing the healthcare costs, with potential applications in other cancer types as well,” said Professor Victoria Mar and Associate Professor Zhaolin Chen.
This project was one of six awarded Monash Institute of Medial Engineering (MIME) Invent Research Support grant support in late 2024.
“The MIME Invent Research Support Grant plays an important role in advancing our work by fostering collaboration between AI experts, imaging scientists, and clinicians. It enables us to accelerate the development of our low-dose PET reconstruction model by providing the resources needed to refine algorithms, curate clinical datasets, and move toward real-world validation.
“This support helps bridge the gap between research and clinical application and allows our team to translate cutting-edge AI into meaningful solutions for melanoma patients,” said Professor Victoria Mar and Associate Professor Zhaolin Chen.
Next steps for the project include validation testing.
Watch this space for further developments.
For more information on MIME Invent Research Support grant support visit: https://www.monash.edu/mime/programs/invent-research-support.
The project team acknowledges the support of the Monash Institute of Medical Engineering (MIME), clinical partners, and the data contributors for enabling this research.