通过双重不变性提升手术阶段识别的自训练方法
Dual Invariance Self-training for Reliable Semi-supervised Surgical Phase Recognition
- 利用时间与变换不变性筛选可靠伪标签
- 在白内障和胆囊切除数据集上超越现有SOTA方法
- 适合缺乏标注数据的医疗视频分析场景
精准的手术阶段识别对推进计算机辅助干预至关重要，但标注数据稀缺限制了深度学习模型的可靠性。半监督学习（SSL）尤其基于伪标签的方法虽有潜力，但常缺乏可靠的伪标签评估机制。为此，我们提出一种新型SSL框架——双重不变性自训练（DIST），融合时间不变性和变换不变性，以增强手术阶段识别能力。该方法采用两步自训练过程，动态选择可信伪标签，确保伪监督的鲁棒性。所提方法有效降低噪声伪标签风险，使决策边界更贴近真实数据分布，提升对未见数据的泛化能力。在白内障（Cataract）和胆囊切除80（Cholec80）数据集上的实验表明，本方法优于当前最先进的SSL方法，在多种网络架构下均持续超越监督与半监督基线。
原文摘要 · Abstract (English)
Accurate surgical phase recognition is crucial for advancing computer-assisted interventions, yet the scarcity of labeled data hinders training reliable deep learning models. Semi-supervised learning (SSL), particularly with pseudo-labeling, shows promise over fully supervised methods but often lacks reliable pseudo-label assessment mechanisms. To address this gap, we propose a novel SSL framework, Dual Invariance Self-Training (DIST), that incorporates both Temporal and Transformation Invariance to enhance surgical phase recognition. Our two-step self-training process dynamically selects reliable pseudo-labels, ensuring robust pseudo-supervision. Our approach mitigates the risk of noisy pseudo-labels, steering decision boundaries toward true data distribution and improving generalization to unseen data. Evaluations on Cataract and Cholec80 datasets show our method outperforms state-of-the-art SSL approaches, consistently surpassing both supervised and SSL baselines across various network architectures.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。