Safeguarding AI in Medical Imaging: Post-Hoc Out-of-Distribution Detection with Normalizing Flows
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866908968418279424 |
|---|---|
| author | Lotfi, Dariush Mahani, Mohammad-Ali Nikouei Koohi-Moghadam, Mohamad Bae, Kyongtae Ty |
| author_facet | Lotfi, Dariush Mahani, Mohammad-Ali Nikouei Koohi-Moghadam, Mohamad Bae, Kyongtae Ty |
| contents | In AI-driven medical imaging, the failure to detect out-of-distribution (OOD) data poses a severe risk to clinical reliability, potentially leading to critical diagnostic errors. Current OOD detection methods often demand impractical retraining or modifications to pre-trained models, hindering their adoption in regulated clinical environments. To address this challenge, we propose a post-hoc normalizing flow-based approach that seamlessly integrates with existing pre-trained models without altering their weights. We evaluate the approach on our in-house-curated MedOOD dataset, designed to capture clinically relevant distribution shifts, and on the MedMNIST benchmark. The proposed method achieves an AUROC of 84.61% on MedOOD, outperforming ViM (80.65%) and MDS (80.87%), and reaches 93.8% AUROC on MedMNIST, surpassing ViM (88.08%) and ReAct (87.05%). This combination of strong performance and post-hoc integration capability makes our approach a practical and effective safeguard for clinical imaging workflows. The model and code to build OOD datasets are publicly accessible at https://github.com/dlotfi/MedOODFlow. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_11638 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Safeguarding AI in Medical Imaging: Post-Hoc Out-of-Distribution Detection with Normalizing Flows Lotfi, Dariush Mahani, Mohammad-Ali Nikouei Koohi-Moghadam, Mohamad Bae, Kyongtae Ty Computer Vision and Pattern Recognition In AI-driven medical imaging, the failure to detect out-of-distribution (OOD) data poses a severe risk to clinical reliability, potentially leading to critical diagnostic errors. Current OOD detection methods often demand impractical retraining or modifications to pre-trained models, hindering their adoption in regulated clinical environments. To address this challenge, we propose a post-hoc normalizing flow-based approach that seamlessly integrates with existing pre-trained models without altering their weights. We evaluate the approach on our in-house-curated MedOOD dataset, designed to capture clinically relevant distribution shifts, and on the MedMNIST benchmark. The proposed method achieves an AUROC of 84.61% on MedOOD, outperforming ViM (80.65%) and MDS (80.87%), and reaches 93.8% AUROC on MedMNIST, surpassing ViM (88.08%) and ReAct (87.05%). This combination of strong performance and post-hoc integration capability makes our approach a practical and effective safeguard for clinical imaging workflows. The model and code to build OOD datasets are publicly accessible at https://github.com/dlotfi/MedOODFlow. |
| title | Safeguarding AI in Medical Imaging: Post-Hoc Out-of-Distribution Detection with Normalizing Flows |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2502.11638 |