Swiss doctors have created a new tool for predicting mesothelioma outcomes with FDG-PET scans.
FDG-PET scanning is a nuclear imaging technique. It gives doctors a non-invasive way to track the growth of pleural mesothelioma tumors.
The Swiss team analyzed multiple PET scans and CT scans from mesothelioma patients. They used a computer to look for commonalities in the scans. Then they used the information to create a computer model for predicting mesothelioma outcomes with FDG-PET.
The model may help doctors and patients make better decisions about mesothelioma treatment in the future.
Imaging Studies for Mesothelioma Prognosis
Pleural mesothelioma tumors occur on the pleural lining around the lungs. In the early stages, mesothelioma causes few symptoms. As mesothelioma tumors grow, they can impact other organs in the chest. If a doctor suspects mesothelioma, he or she may order an imaging scan.
Imaging studies like CT and FDG-PET help doctors diagnose and predict mesothelioma outcomes. These tools make it possible to look inside the chest without making incisions. They provide information about the size and location of tumors which can impact mesothelioma prognosis.
CT scans take X-rays from different angles. The X-rays get compiled into a single 3D image. Predicting mesothelioma outcomes with FDG-PET relies on a tracer. Cancer cells absorb more of the FDG tracer so they show up brighter on the PET image.
But these scans can be tricky to read. Unless a doctor is very experienced at predicting mesothelioma outcomes with FDG-PET or CT, they may miss some signs of mesothelioma.
A Better Tool for Predicting Mesothelioma Outcomes with FDG-PET?
Machine learning is a way of teaching computers to find patterns in data. The Swiss team hoped it could improve accuracy in reading PET and CT scans for mesothelioma prognosis.
They collected imaging scans from 72 mesothelioma patients. The patients had surgery at their hospital between 2009 and 2017.
A computer program analyzed the mesothelioma imaging scans. It identified more than 1,400 features from each image. The team picked out the features that were most closely linked to mesothelioma prognosis. They used these features to “teach” the computer to predict mesothelioma outcomes. This modeling is called radiomics.
The computer could not create an accurate radiomics model from the CT data. But the FDG-PET images worked better. The computer model made from these images proved to be an accurate way of predicting mesothelioma outcomes with FDG-PET.
“We were able to build a successful FDG-PET radiomics model for the prediction of progression free survival in malignant pleural mesothelioma,” say the Swiss researchers. “Radiomics could serve as a tool to aid clinical decision support systems for treatment of malignant pleural mesothelioma in future.”
Pavic, M, et al, “FDG PET versus CT radiomics to predict outcome in malignant pleural mesothelioma patients”, July 13, 2020, EJNMMI Research, https://ejnmmires.springeropen.com/articles/10.1186/s13550-020-00669-3