MedFormer-UR: Uncertainty-Routed Transformer for Medical Image Classification
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866918438111281152 |
|---|---|
| author | Sibhai, Mohammed Maaz Alkhateeb, Abedalrhman Ahmed, Saad B. |
| author_facet | Sibhai, Mohammed Maaz Alkhateeb, Abedalrhman Ahmed, Saad B. |
| contents | To ensure safe clinical integration, deep learning models must provide more than just high accuracy; they require dependable uncertainty quantification. While current Medical Vision Transformers perform well, they frequently struggle with overconfident predictions and a lack of transparency, issues that are magnified by the noisy and imbalanced nature of clinical data. To address this, we enhanced the modified Medical Transformer (MedFormer) that incorporates prototype-based learning and uncertainty-guided routing, by utilizing a Dirichlet distribution for per-token evidential uncertainty, our framework can quantify and localize ambiguity in real-time. This uncertainty is not just an output but an active participant in the training process, filtering out unreliable feature updates. Furthermore, the use of class-specific prototypes ensures the embedding space remains structured, allowing for decisions based on visual similarity. Testing across four modalities (mammography, ultrasound, MRI, and histopathology) confirms that our approach significantly enhances model calibration, reducing expected calibration error (ECE) by up to 35%, and improves selective prediction, even when accuracy gains are modest. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_08868 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | MedFormer-UR: Uncertainty-Routed Transformer for Medical Image Classification Sibhai, Mohammed Maaz Alkhateeb, Abedalrhman Ahmed, Saad B. Image and Video Processing Artificial Intelligence Computer Vision and Pattern Recognition Machine Learning To ensure safe clinical integration, deep learning models must provide more than just high accuracy; they require dependable uncertainty quantification. While current Medical Vision Transformers perform well, they frequently struggle with overconfident predictions and a lack of transparency, issues that are magnified by the noisy and imbalanced nature of clinical data. To address this, we enhanced the modified Medical Transformer (MedFormer) that incorporates prototype-based learning and uncertainty-guided routing, by utilizing a Dirichlet distribution for per-token evidential uncertainty, our framework can quantify and localize ambiguity in real-time. This uncertainty is not just an output but an active participant in the training process, filtering out unreliable feature updates. Furthermore, the use of class-specific prototypes ensures the embedding space remains structured, allowing for decisions based on visual similarity. Testing across four modalities (mammography, ultrasound, MRI, and histopathology) confirms that our approach significantly enhances model calibration, reducing expected calibration error (ECE) by up to 35%, and improves selective prediction, even when accuracy gains are modest. |
| title | MedFormer-UR: Uncertainty-Routed Transformer for Medical Image Classification |
| topic | Image and Video Processing Artificial Intelligence Computer Vision and Pattern Recognition Machine Learning |
| url | https://arxiv.org/abs/2604.08868 |