# DINOv3-MIL: Per-Kidney Multi-Label Tumour and Cyst Detection from Foundation-Model Patch Tokens on KiTS23

arXiv: https://arxiv.org/abs/2607.22687 · HTML: https://mlsanity.com/paper/2607.22687/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 4 of its 7 references that are in the map.
- First posted on arXiv 16 Jul 2026.

## Builds on (most cited first, 4 of 4)

- [Sanity Checks for Saliency Maps](https://mlsanity.com/paper/1810.03292/) (2018)
- [Attention-based Deep Multiple Instance Learning](https://mlsanity.com/paper/1802.04712/) (2018)
- [This Looks Like That: Deep Learning for Interpretable Image Recognition](https://mlsanity.com/paper/1806.10574/) (2018)
- [Interpretable Image Classification with Adaptive Prototype-based Vision Transformers](https://mlsanity.com/paper/2410.20722/) (2024)

## Abstract (arXiv)

Foundation vision models trained on natural images transfer to medical tasks without domain pre-training, but volumetric classification requires aggregating tens of thousands of patch tokens per study, and the aggregator constrains how the resulting model can be interpreted. We compare three aggregators on identical frozen DINOv3 ViT-H/16+ features for renal tumour/cyst detection on KiTS23 (966 kidneys; n=97 test): a CLS-token linear probe, gated attention multiple instance learning (MIL) over 55,296 patch tokens, and a prototype head following ProtoViT. Attention MIL achieves the highest AUROC for tumour (0.74, 95% CI 0.64-0.83) and cyst (0.80, 0.70-0.88), with attention enriched 7.5-9.8x over chance within annotated lesions. The prototype head does not transfer to cyst detection (AUROC 0.51), exposing an interpretability-performance trade-off at this token scale.