ACQI and UWA Medical Physics Group PhD student Jake Kendrick has just published the results of his study into automated detection and segmentation of prostate cancer metastases in PSMA PET images. Jake collaborated with a team of physicists, computer scientists and nuclear medicine specialists to use deep learning to replicate the image assessment of an expert human observer and produce a prognostic model with similar accuracy. This is paving the way for rapid, automated assessment of these images to enable better understanding of the disease and a more patient-specific process for treatment decision making. The paper is accessible here.
Australian Centre for Quantitative Imaging 