计算机科学
分割
人工智能
稳健性(进化)
一般化
注释
模式识别(心理学)
一致性(知识库)
监督学习
投影(关系代数)
机器学习
数学
算法
人工神经网络
数学分析
生物化学
化学
基因
作者
Feng Gao,Minhao Hu,Min-Er Zhong,Feng Su,Xuwei Tian,Xiaochun Meng,Mayidili Nijiati,Zhichao Huang,Min‐Yi Lv,Tao Song,Xiaofan Zhang,Xiaoguang Zou,Xiaojian Wu
标识
DOI:10.1016/j.media.2022.102515
摘要
Since segmentation labeling is usually time-consuming and annotating medical images requires professional expertise, it is laborious to obtain a large-scale, high-quality annotated segmentation dataset. We propose a novel weakly- and semi-supervised framework named SOUSA (Segmentation Only Uses Sparse Annotations), aiming at learning from a small set of sparse annotated data and a large amount of unlabeled data. The proposed framework contains a teacher model and a student model. The student model is weakly supervised by scribbles and a Geodesic distance map derived from scribbles. Meanwhile, a large amount of unlabeled data with various perturbations are fed to student and teacher models. The consistency of their output predictions is imposed by Mean Square Error (MSE) loss and a carefully designed Multi-angle Projection Reconstruction (MPR) loss. Extensive experiments are conducted to demonstrate the robustness and generalization ability of our proposed method. Results show that our method outperforms weakly- and semi-supervised state-of-the-art methods on multiple datasets. Furthermore, our method achieves a competitive performance with some fully supervised methods with dense annotation when the size of the dataset is limited.
科研通智能强力驱动
Strongly Powered by AbleSci AI