Publications by authors named "Seohoon Lim"
Despite the fact that digital pathology has provided a new paradigm for modern medicine, the insufficiency of annotations for training remains a significant challenge. Due to the weak generalization abilities of deep-learning models, their performance is notably constrained in domains without sufficient annotations. Our research aims to enhance the model's generalization ability through domain adaptation, increasing the prediction ability for the target domain data while only using the source domain labels for training.
View Article and Find Full Text PDFArticle Synopsis
- Deep-learning-based survival prediction helps doctors assess patients' death risk and estimate survival times, utilizing techniques like the Cox model.
- The paper introduces a new method that merges risk and survival time predictions by using features from risk predictions to improve accuracy in forecasting survival time.
- It employs high-resolution whole slide images (WSIs) to extract tumor patches and uses a graph convolutional network to enhance information aggregation, leading to significantly better prediction accuracy compared to existing methods.