发表机构
Southern University of Science and Technology; Zhejiang University; Tianjin University（南方科技大学; 浙江大学; 天津大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对自推进内窥镜机器人，提出ColoACT系统，采用RGB-D-E动作分块Transformer策略，通过多线索增强和时间集成实现平滑自主结肠导航，在离体猪结肠中达到85.4%和72.5%的直段和弯曲段成功率。
AI 中文摘要
自主结肠镜导航可以减少操作者负担以及环形成或组织创伤的风险，但由于可变形解剖结构、弱纹理和高光内镜视觉以及富含接触的粘弹性相互作用，仍然具有挑战性。现有方法要么依赖几何驱动的流程，这些流程高效且可解释，但由于手工设计的特征和切换逻辑而脆弱；要么采用基于学习的策略，其推断的深度/几何在弱纹理和高光下可能变得时间上不一致或过度平滑，而模拟训练的变体（例如深度强化学习）可能进一步遭受模拟到现实的差距。我们提出了ColoACT，一种自主导航系统，集成了基于RGB-D-E的动作分块Transformer策略（ColoACT策略），用于紧凑型自推进斜齿轮内窥镜机器人（BGER）。ColoACT策略通过估计的相对深度和基于梯度的伪高程图增强RGB，以增强褶皱脊显著性和其他高频几何线索，并通过预测重叠动作块并经由时间集成融合，实现对BGER的平滑连续控制。在不同的离体猪结肠（约60厘米）中，我们的系统在直段和弯曲段分别实现了85.4%和72.5%的成功率，在90度转弯中实现70%的成功率，在双弯序列中实现60%的成功率，并在具有挑战性的三弯段中进一步证明了可行性。项目页面可在以下网址获取：this https URL。
英文摘要
Autonomous colonoscopic navigation can reduce operator burden and the risk of loop formation or tissue trauma, but remains challenging due to deformable anatomy, weak-texture and specular endoscopic visuals, and contact-rich viscoelastic interactions. Existing methods either rely on geometry-driven pipelines, which are efficient and interpretable yet brittle due to manually engineered features and switching logic, or adopt learning-based policies, whose inferred depth/geometry can become temporally inconsistent or overly smooth under weak texture and specular highlights while simulation-trained variants (e.g., deep reinforcement learning) may further suffer from a sim-to-real gap. We propose ColoACT, an autonomous navigation system that integrates an RGB-D-E based Action Chunking Transformer policy (ColoACT policy) for a compact self-propelled Bevel-Gear-Based Endoscopic Robot (BGER). The ColoACT policy augments RGB with estimated relative depth and a gradient-based pseudo-elevation map to enhance fold-ridge saliency and other high-frequency geometric cues, and enables smooth continuous control of the BGER by predicting overlapping action chunks and fusing them via temporal ensembling. In different \textit{ex-vivo} porcine colons (approximately 60 cm), our system achieves success rates of 85.4\% and 72.5\% in straight and curved segments, respectively, and achieves 70\% success in 90-degree turns and 60\% in double-bend sequences, with feasibility further demonstrated in challenging triple-bend segments. The project page is available at: https://Adamhu1.github.io/ColoACT/.
