# Med-Banana: Learning Quality-Controlled Medical Image Editing from Success-and-Failure Trajectories

arXiv: https://arxiv.org/abs/2511.00801 · HTML: https://mlsanity.com/paper/2511.00801/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 10 of its 19 references that are in the map.
- First posted on arXiv 2 Nov 2025.

## Builds on (most cited first, 10 of 10)

- [The Unreasonable Effectiveness of Deep Features as a Perceptual Metric](https://mlsanity.com/paper/1801.03924/) (2018)
- [MLLM-as-a-Judge: Assessing Multimodal LLM-as-a-Judge with Vision-Language Benchmark](https://mlsanity.com/paper/2402.04788/) (2024)
- [Revisiting Classifier Two-Sample Tests](https://mlsanity.com/paper/1610.06545/) (2016)
- [A Benchmark of Ocular Disease Intelligent Recognition: One Shot for Multi-disease Detection](https://mlsanity.com/paper/2102.07978/) (2021)
- [PRISM: High-Resolution & Precise Counterfactual Medical Image Generation using Language-guided Stable Diffusion](https://mlsanity.com/paper/2503.00196/) (2025)
- [Latent Drifting in Diffusion Models for Counterfactual Medical Image Synthesis](https://mlsanity.com/paper/2412.20651/) (2024)
- [Generation of Structurally Realistic Retinal Fundus Images with Diffusion Models](https://mlsanity.com/paper/2305.06813/) (2023)
- [MedEdit: Counterfactual Diffusion-based Image Editing on Brain MRI](https://mlsanity.com/paper/2407.15270/) (2024)
- [MedEBench: Diagnosing Reliability in Text-Guided Medical Image Editing](https://mlsanity.com/paper/2506.01921/) (2025)
- [An LLM-LVLM Driven Agent for Iterative and Fine-Grained Image Editing](https://mlsanity.com/paper/2508.17435/) (2025)

## Built on by (most cited first, 1 of 1)

- [MieDB-100k: A Comprehensive Dataset for Medical Image Editing](https://mlsanity.com/paper/2602.09587/) (2026)

## Abstract (arXiv)

Text-guided medical image editing must satisfy the requested pathology while preserving anatomy, modality-specific appearance, and clinical plausibility. However, existing datasets largely supervise editors with final accepted edits and discard the failed attempts produced during generation. We argue that these failures provide essential supervision for quality control: they specify what should be rejected, why an edit is medically or visually invalid, and how the instruction should be revised. We present Med-Banana, a trajectory-supervised framework for quality-controlled medical image editing. We introduce Med-Banana-80K, a large-scale resource of success-and-failure editing trajectories with candidate images, verification outcomes, rejection reasons, and prompt refinements. Building on it, Med-Banana jointly trains an editor, verifier, and refiner, enabling edit--verify--refine inference from accepted and rejected attempts. Experiments across MLLM judges, blind expert assessment, source-preservation and real--synthetic separability probes demonstrate consistent improvements over open medical image editors. Code and data are publicly available.