Two Papers Accepted at MICCAI 2023
Published:
ContourAId’s foundational research continues to gain recognition with two papers accepted at MICCAI 2023 in Vancouver, one of the premier conferences in medical image computing.
Our paper “Do we really need that skip connection?” was accepted as an Early Accept (top 14% of submissions), examining architectural choices in deep learning models for medical image segmentation. This work directly informs the design of ContourAId’s auto-contouring algorithms, ensuring optimal performance across diverse imaging scenarios.
The second paper, “Dose Guidance for Radiotherapy-oriented Deep Learning Segmentation,” introduces novel methods for incorporating dosimetric information into AI-based contouring. This research forms the basis of ContourAId’s unique approach to quality-aware segmentation, where AI models are trained to prioritize clinically relevant accuracy.
These publications represent significant contributions to the medical imaging community and demonstrate the rigorous scientific foundation underlying ContourAId’s technology.