Dr Barnes sat down with the PACT team to talk about her project Discovering Radiological Biomarkers in Silicosis using Artificial Intelligence, including how it started, and what the support of the PACT Grant-in-Aid will mean for this research.
Tell us a bit about yourself and why you chose PF as your research area:
I am a Respiratory Physician (Alfred Health) and have completed an overseas fellowship in ILD (UCSF) and PhD (Monash University). I was drawn to PF research because seeing patients in clinic I realised there are many gaps in our knowledge about the best methods of diagnosis and treatment, and wanted to provide a better experience for patients. I was also drawn to this field because there are many clinician-scientists who I have worked with who are inspiring and collaborative and great to work with!
Describe your clinical research project in a couple of sentences:
I am currently working on artificial intelligence biomarkers to better analyse CT scans for silicosis patients, to predict who will develop fibrosis. Unfortunately there has been a surge in silicosis cases in Australia from those working in the stone benchtop industry. Screening programs are now in place, but it is difficult to determine who will progress to Pulmonary Fibrosis. Developing biomarkers using AI means these biomarkers can be used in settings where expert Radiologists are not so accessible.
Tell us how this project came about:
This work extends from my PhD work which examined AI CT biomarkers in hypersensitivity pneumonitis. There are lots of exciting projects using AI in PF, and several researchers in the PACT network including Prof Corte and Dr John Macintosh doing great research in this area.
What do you hope this project will achieve?
We hope this project will identify biomarkers that can be used in detecting which patients with silicosis will develop Pulmonary Fibrosis. We also expect this technology can be translated to use in other Pulmonary Fibrosis conditions.
What has been the biggest challenge of the project so far?
The data is never clean! It often requires a lot of pre-processing before the analysis can occur – which always takes longer than planned!
The people working on this project are:
I am working on this project with Dr Ryan Hoy, who is a renowned expert in silicosis, and who set up the Monash Silicosis Registry, and collaborating with AI experts from the University of Michigan.
What will the support of the PACT Grant-in-Aid mean for your clinical research?
The PACT Grant-in-aid will support the expertise of AI engineers to get the best out of our data.
What do you see as the biggest evidence gaps in PF care? What are the clinical research priorities?
I think so much work has gone into IPF which has greatly advanced our knowledge and options for treatment, but I think there is still much work to do with non-IPF patients with PF, such as hypersensitivity pneumonitis and silicosis. Identifying biomarkers of disease behaviour will help determine what additional treatments might be useful in these patients.