# Attention Consistent Longitudinal Medical Visual Question Answering Guided by Vision Foundation Models

arXiv: https://arxiv.org/abs/2606.06534 · HTML: https://mlsanity.com/paper/2606.06534/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 19 of its 42 references that are in the map.
- First posted on arXiv 3 Jun 2026.

## Builds on (most cited first, 19 of 19)

- [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://mlsanity.com/paper/2103.14030/) (2021)
- [Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization](https://mlsanity.com/paper/1610.02391/) (2016)
- [SAM 2: Segment Anything in Images and Videos](https://mlsanity.com/paper/2408.00714/) (2024)
- [TransUNet: Transformers Make Strong Encoders for Medical Image Segmentation](https://mlsanity.com/paper/2102.04306/) (2021)
- [LLaVA-Med: Training a Large Language-and-Vision Assistant for Biomedicine in One Day](https://mlsanity.com/paper/2306.00890/) (2023)
- [ImageNet-21K Pretraining for the Masses](https://mlsanity.com/paper/2104.10972/) (2021)
- [MIMIC-CXR-JPG, a large publicly available database of labeled chest radiographs](https://mlsanity.com/paper/1901.07042/) (2019)
- [PMC-VQA: Visual Instruction Tuning for Medical Visual Question Answering](https://mlsanity.com/paper/2305.10415/) (2023)
- [Exploring Visual Relationship for Image Captioning](https://mlsanity.com/paper/1809.07041/) (2018)
- [Exploring scalable medical image encoders beyond text supervision](https://mlsanity.com/paper/2401.10815/) (2024)
- [Towards a clinically accessible radiology foundation model: open-access and lightweight, with automated evaluation](https://mlsanity.com/paper/2403.08002/) (2024)
- [Describing and Localizing Multiple Changes with Transformers](https://mlsanity.com/paper/2103.14146/) (2021)
- [Expert Knowledge-Aware Image Difference Graph Representation Learning for Difference-Aware Medical Visual Question Answering](https://mlsanity.com/paper/2307.11986/) (2023)
- [SegDINO: An Efficient Design for Medical and Natural Image Segmentation with DINO-V3](https://mlsanity.com/paper/2509.00833/) (2025)
- [Attention-based Saliency Maps Improve Interpretability of Pneumothorax Classification](https://mlsanity.com/paper/2303.01871/) (2023)
- [One Map Does Not Fit All: Evaluating Saliency Map Explanation on Multi-Modal Medical Images](https://mlsanity.com/paper/2107.05047/) (2021)
- [Co-Saliency Detection with Co-Attention Fully Convolutional Network](https://mlsanity.com/paper/2008.08909/) (2020)
- [Pretraining Vision-Language Model for Difference Visual Question Answering in Longitudinal Chest X-rays](https://mlsanity.com/paper/2402.08966/) (2024)
- [Comparing radiologists' gaze and saliency maps generated by interpretability methods for chest x-rays](https://mlsanity.com/paper/2112.11716/) (2021)

## Abstract (arXiv)

Longitudinal medical visual question answering (VQA) requires reasoning about anatomical differences between an image of a current time point and an image of a referred time point. We propose an attention-guided encoder-decoder for this task with chest X-rays. Instead of conventional direct contrast, we propose to include a lightweight affine registration module to reduce nuisance motion by co-registering the current image to the reference image with a small registration regularizer. The registered image pair is fed into the image encoder, followed by a frozen DINO-based mask generator and a trainable adaptive mask generator to produce masks applied to the original image pairs. The masked image pairs are again fed into the image encoder and concatenated with text features as the input to a multimodal transformer-based decoder to generate final answers. To facilitate learning stabilization and clarify the change signal, inspired by DINO-v3, we include additional auxiliary objectives, including a mask rebuilding loss, a pairwise Gram-style consistency loss, and a KoLeo uniformity loss, which enhances the geometry of the representation. On the Medical-Diff-VQA benchmark, the model delivers strong BLEU, ROUGE-L, CIDEr, and METEOR scores while offering intrinsic interpretability through the shared saliency mask. These results support saliency-conditioned generation with mild pre-alignment as a principled framework for longitudinal reasoning in medical VQA. Our training strategy also illustrates the potential of a paradigm in utilizing image foundation models in biomedicine: optimizing both supervised and unsupervised learning objectives simultaneously.