# BleedOrigin: Dynamic Bleeding Source Localization in Endoscopic Submucosal Dissection via Dual-Stage Detection and Tracking

arXiv: https://arxiv.org/abs/2507.15094 · HTML: https://mlsanity.com/paper/2507.15094/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 33 of its 58 references that are in the map.
- First posted on arXiv 20 Jul 2025.

## Builds on (most cited first, 30 of 33)

- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks](https://mlsanity.com/paper/1506.01497/) (2015)
- [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023)
- [Generative Adversarial Networks](https://mlsanity.com/paper/2203.00667/) (2022)
- [You Only Look Once: Unified, Real-Time Object Detection](https://mlsanity.com/paper/1506.02640/) (2015)
- [CoTracker: It is Better to Track Together](https://mlsanity.com/paper/2307.07635/) (2023)
- [YOLOv11: An Overview of the Key Architectural Enhancements](https://mlsanity.com/paper/2410.17725/) (2024)
- [AdaLoRA: Adaptive Budget Allocation for Parameter-Efficient Fine-Tuning](https://mlsanity.com/paper/2303.10512/) (2023)
- [Perception Encoder: The best visual embeddings are not at the output of the network](https://mlsanity.com/paper/2504.13181/) (2025)
- [TAPIR: Tracking Any Point with per-frame Initialization and temporal Refinement](https://mlsanity.com/paper/2306.08637/) (2023)
- [BootsTAP: Bootstrapped Training for Tracking-Any-Point](https://mlsanity.com/paper/2402.00847/) (2024)
- [XFeat: Accelerated Features for Lightweight Image Matching](https://mlsanity.com/paper/2404.19174/) (2024)
- [TAPVid-3D: A Benchmark for Tracking Any Point in 3D](https://mlsanity.com/paper/2407.05921/) (2024)
- [MFT: Long-Term Tracking of Every Pixel](https://mlsanity.com/paper/2305.12998/) (2023)
- [Parameter-Efficient Fine-Tuning for Medical Image Analysis: The Missed Opportunity](https://mlsanity.com/paper/2305.08252/) (2023)
- [Video Object Detection with an Aligned Spatial-Temporal Memory](https://mlsanity.com/paper/1712.06317/) (2017)
- [EndoChat: Grounded Multimodal Large Language Model for Endoscopic Surgery](https://mlsanity.com/paper/2501.11347/) (2025)
- [Looking Fast and Slow: Memory-Guided Mobile Video Object Detection](https://mlsanity.com/paper/1903.10172/) (2019)
- [Track-On: Transformer-based Online Point Tracking with Memory](https://mlsanity.com/paper/2501.18487/) (2025)
- [Surgical Tattoos in Infrared: A Dataset for Quantifying Tissue Tracking and Mapping](https://mlsanity.com/paper/2309.16782/) (2023)
- [TEMPURA: Temporal Event Masked Prediction and Understanding for Reasoning in Action](https://mlsanity.com/paper/2505.01583/) (2025)
- [Autonomous Robotic Suction to Clear the Surgical Field for Hemostasis using Image-based Blood Flow Detection](https://mlsanity.com/paper/2010.08441/) (2020)
- [PACE: A Large-Scale Dataset with Pose Annotations in Cluttered Environments](https://mlsanity.com/paper/2312.15130/) (2023)
- [Motion2Vec: Semi-Supervised Representation Learning from Surgical Videos](https://mlsanity.com/paper/2006.00545/) (2020)
- [Vision Foundation Models in Medical Image Analysis: Advances and Challenges](https://mlsanity.com/paper/2502.14584/) (2025)
- [Online Temporal Action Localization with Memory-Augmented Transformer](https://mlsanity.com/paper/2408.02957/) (2024)
- [MFTIQ: Multi-Flow Tracker with Independent Matching Quality Estimation](https://mlsanity.com/paper/2411.09551/) (2024)
- [Adapting Vision-Language Foundation Model for Next Generation Medical Ultrasound Image Analysis](https://mlsanity.com/paper/2506.08849/) (2025)
- [WCEbleedGen: A wireless capsule endoscopy dataset and its benchmarking for automatic bleeding classification, detection, and segmentation](https://mlsanity.com/paper/2408.12466/) (2024)
- [HemoSet: The First Blood Segmentation Dataset for Automation of Hemostasis Management](https://mlsanity.com/paper/2403.16286/) (2024)

## Built on by (most cited first, 2 of 2)

- [MI-CXR: A Benchmark for Longitudinal Reasoning over Multi-Interval Chest X-rays](https://mlsanity.com/paper/2605.15574/) (2026)
- [DBT-Bleed: Dual-Branch Temporal Modeling with Key-Frame Selection for Surgical Bleeding Detection](https://mlsanity.com/paper/2606.22829/) (2026)

## Abstract (arXiv)

Intraoperative bleeding during Endoscopic Submucosal Dissection (ESD) poses significant risks, demanding precise, real-time localization and continuous monitoring of the bleeding source for effective hemostatic intervention. In particular, endoscopists have to repeatedly flush to clear blood, allowing only milliseconds to identify bleeding sources, an inefficient process that prolongs operations and elevates patient risks. However, current Artificial Intelligence (AI) methods primarily focus on bleeding region segmentation, overlooking the critical need for accurate bleeding source detection and temporal tracking in the challenging ESD environment, which is marked by frequent visual obstructions and dynamic scene changes. This gap is widened by the lack of specialized datasets, hindering the development of robust AI-assisted guidance systems. To address these challenges, we introduce BleedOrigin-Bench, the first comprehensive ESD bleeding source dataset, featuring 1,771 expert-annotated bleeding sources across 106,222 frames from 44 procedures, supplemented with 39,755 pseudo-labeled frames. This benchmark covers 8 anatomical sites and 6 challenging clinical scenarios. We also present BleedOrigin-Net, a novel dual-stage detection-tracking framework for the bleeding source localization in ESD procedures, addressing the complete workflow from bleeding onset detection to continuous spatial tracking. We compare with widely-used object detection models (YOLOv11/v12), multimodal large language models, and point tracking methods. Extensive evaluation demonstrates state-of-the-art performance, achieving 96.85% frame-level accuracy (±≤8 frames) for bleeding onset detection, 70.24% pixel-level accuracy (≤100 px) for initial source detection, and 96.11% pixel-level accuracy (≤100 px) for point tracking.