Douwe has developed a minimally interactive deep learning-based segmentation method for soft-tissue tumors (STTs) on CT and MRI, and his considerable contribution were outlined in the paper, “Minimally interactive segmentation of soft-tissue tumors on CT and MRI using deep learning”
Segmentations are crucial in medical imaging for morphological, volumetric, and radiomics biomarkers. Manual segmentation is accurate but not feasible in clinical workflow, while automatic segmentation generally performs sub-par.
In this work, they (Douwe with the co-authors, in particular, Martijn, Wiro, and Stefan from the BIGR) present a deep learning-based segmentation method that requires minimal user interaction, drastically reducing the time needed to produce accurate segmentations (See above figure presenting schematic overview). Their approach not only outperforms existing automatic segmentation methods—often challenged by the diverse and heterogeneous nature of soft-tissue tumors—but also builds upon and improves earlier interactive segmentation methods.
They validated the proposed method for both volumetric measurements and radiomics analysis, demonstrating its potential for clinical use.
Read a related post here and the full open-access article here.