A self-training teacher-student model with an automatic label grader for abdominal skeletal muscle segmentation

作者：

Highlights：

• We designed a semi-supervised learning pipeline for skeletal muscle segmentation.

• An automatic label grader was proposed for quality control of pseudo labels.

• Our method starts with a small, labeled dataset and leverages large unlabeled data.

• We performed extensive experiments and our method outperformed other methods.

摘要

•We designed a semi-supervised learning pipeline for skeletal muscle segmentation.•An automatic label grader was proposed for quality control of pseudo labels.•Our method starts with a small, labeled dataset and leverages large unlabeled data.•We performed extensive experiments and our method outperformed other methods.

论文关键词：Image segmentation,Semi-supervised learning,Self-attention,Teacher-student model,Skeletal muscle

论文评审过程：Received 1 July 2021, Revised 15 June 2022, Accepted 14 July 2022, Available online 19 July 2022, Version of Record 30 July 2022.

论文官网地址：https://doi.org/10.1016/j.artmed.2022.102366