# Divide-then-Diagnose: Weaving Clinician-Inspired Contexts for Ultra-Long Capsule Endoscopy Videos

arXiv: https://arxiv.org/abs/2604.21814 · HTML: https://mlsanity.com/paper/2604.21814/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 21 of its 53 references that are in the map.
- First posted on arXiv 23 Apr 2026.

## Builds on (most cited first, 21 of 21)

- [Qwen2.5-VL Technical Report](https://mlsanity.com/paper/2502.13923/) (2025)
- [Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution](https://mlsanity.com/paper/2409.12191/) (2024)
- [Qwen3-VL Technical Report](https://mlsanity.com/paper/2511.21631/) (2025)
- [InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models](https://mlsanity.com/paper/2504.10479/) (2025)
- [InternVL3.5: Advancing Open-Source Multimodal Models in Versatility, Reasoning, and Efficiency](https://mlsanity.com/paper/2508.18265/) (2025)
- [DINOv3](https://mlsanity.com/paper/2508.10104/) (2025)
- [MedGemma Technical Report](https://mlsanity.com/paper/2507.05201/) (2025)
- [Lingshu: A Generalist Foundation Model for Unified Multimodal Medical Understanding and Reasoning](https://mlsanity.com/paper/2506.07044/) (2025)
- [HuatuoGPT-Vision, Towards Injecting Medical Visual Knowledge into Multimodal LLMs at Scale](https://mlsanity.com/paper/2406.19280/) (2024)
- [Scaling Video-Language Models to 10K Frames via Hierarchical Differential Distillation](https://mlsanity.com/paper/2504.02438/) (2025)
- [LongVideoAgent: Multi-Agent Reasoning with Long Videos](https://mlsanity.com/paper/2512.20618/) (2025)
- [Frontiers in Intelligent Colonoscopy](https://mlsanity.com/paper/2410.17241/) (2024)
- [EndoL2H: Deep Super-Resolution for Capsule Endoscopy](https://mlsanity.com/paper/2002.05459/) (2020)
- [EndoOOD: Uncertainty-aware Out-of-distribution Detection in Capsule Endoscopy Diagnosis](https://mlsanity.com/paper/2402.11476/) (2024)
- [Capsule Endoscopy Multi-classification via Gated Attention and Wavelet Transformations](https://mlsanity.com/paper/2410.19363/) (2024)
- [Domain-Adaptive Pre-training of Self-Supervised Foundation Models for Medical Image Classification in Gastrointestinal Endoscopy](https://mlsanity.com/paper/2410.21302/) (2024)
- [CAVE-Net: Classifying Abnormalities in Video Capsule Endoscopy](https://mlsanity.com/paper/2410.20231/) (2024)
- [Influence of color correction on pathology detection in Capsule Endoscopy](https://mlsanity.com/paper/2502.00076/) (2025)
- [ColonNet: A Hybrid Of DenseNet121 And U-NET Model For Detection And Segmentation Of GI Bleeding](https://mlsanity.com/paper/2412.05216/) (2024)
- [Transformer-Based Wireless Capsule Endoscopy Bleeding Tissue Detection and Classification](https://mlsanity.com/paper/2412.19218/) (2024)
- [Colon-X: Advancing Intelligent Colonoscopy toward Clinical Reasoning](https://mlsanity.com/paper/2512.03667/) (2025)

## Built on by (most cited first, 3 of 3)

- [CLIR-Bench: Benchmarking Multimodal Question Answering over Irregular Clinical Time Series](https://mlsanity.com/paper/2607.09880/) (2026)
- [MedHorizon: Towards Long-context Medical Video Understanding in the Wild](https://mlsanity.com/paper/2605.06537/) (2026)
- [Beyond Frame Selection: Generative Latent Evidence Aggregation for Long-Video Understanding](https://mlsanity.com/paper/2607.28516/) (2026)

## Abstract (arXiv)

Capsule endoscopy (CE) enables non-invasive gastrointestinal screening, but current CE research remains largely limited to frame-level classification and detection, leaving video-level analysis underexplored. To bridge this gap, we introduce and formally define a new task, diagnosis-driven CE video summarization, which requires extracting key evidence frames that covers clinically meaningful findings and making accurate diagnoses from those evidence frames. This setting is challenging because diagnostically relevant events are extremely sparse and can be overwhelmed by tens of thousands of redundant normal frames, while individual observations are often ambiguous due to motion blur, debris, specular highlights, and rapid viewpoint changes. To facilitate research in this direction, we introduce VideoCAP, the first CE dataset with diagnosis-driven annotations derived from real clinical reports. VideoCAP comprises 240 full-length videos and provides realistic supervision for both key evidence frame extraction and diagnosis. To address this task, we further propose DiCE, a clinician-inspired framework that mirrors the standard CE reading workflow. DiCE first performs efficient candidate screening over the raw video, then uses a Context Weaver to organize candidates into coherent diagnostic contexts that preserve distinct lesion events, and an Evidence Converger to aggregate multi-frame evidence within each context into robust clip-level judgments. Experiments show that DiCE consistently outperforms state-of-the-art methods, producing concise and clinically reliable diagnostic summaries. These results highlight diagnosis-driven contextual reasoning as a promising paradigm for ultra-long CE video summarization.