The MICCAI 2024 conference is set to take place in Marrakech, Morocco, in October 2024. On June 18th, a research paper titled "Pathological Semantics-Preserving Learning for H&E-to-IHC Virtual Staining" with Dr. Wenjian Qin as the corresponding author and student Fuqiang Chen as the first author, was accepted by MICCAI 2024!
MICCAI stands for International Conference on Medical Image Computing and Computer Assisted Intervention, which is a top-tier international conference in the field of medical image analysis. MICCAI is dedicated to promoting, safeguarding, and advancing research, education, and practice in the field of medical image computing and computer-assisted medical intervention, including biomedical imaging and medical robotics. The association facilitates and fosters the exchange and dissemination of advanced knowledge, professional techniques, and experiences generated by leading institutions and distinguished scientists, doctors, and educators in this field through the organization and hosting of high-quality annual international conferences, workshops, tutorials, and publications.
In "Pathological Semantics-Preserving Learning for H&E-to-IHC Virtual Staining," we introduced two learning strategies: Protein-Aware Learning Strategy (PALS) and Prototype-Consistent Learning Strategy (PCLS), to address the insufficient extraction of pathological semantic information and the misalignment of pathological semantic space in the H&E to IHC virtual staining task. This method accurately quantifies the expression of tumors in label images to constrain key semantic information and uses prototype learning to align the morphology, color, and other features of generated tumor cells semantically with real tumor cells. Additionally, the introduction of focal optical density value assessment in the quantification of tumor cell expression allows the model to focus more on tumor areas, thereby enhancing model performance.
The code is available: https://github.com/ccitachi/PSPStai .
The paper is now available at: https://arxiv.org/abs/2407.03655 .
