Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | https://arxiv.org/abs/2603.21809 |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866910065435344896 |
|---|---|
| author | Imans, Dillan Bui, Phuoc-Nguyen Le, Duc-Tai Choo, Hyunseung |
| author_facet | Imans, Dillan Bui, Phuoc-Nguyen Le, Duc-Tai Choo, Hyunseung |
| contents | Retinal fundus imaging enables low-cost and scalable hypertension (HTN) screening, but HTN-related retinal cues are subtle, yielding high-variance predictions. Brain MRI provides stronger vascular and small-vessel-disease markers of HTN, yet it is expensive and rarely acquired alongside fundus images, resulting in modality-siloed datasets with disjoint MRI and fundus cohorts. We study this unpaired MRI-fundus regime and introduce Clinical Graph-Mediated Distillation (CGMD), a framework that transfers MRI-derived HTN knowledge to a fundus model without paired multimodal data. CGMD leverages shared structured biomarkers as a bridge by constructing a clinical similarity kNN graph spanning both cohorts. We train an MRI teacher, propagate its representations over the graph, and impute brain-informed representation targets for fundus patients. A fundus student is then trained with a joint objective combining HTN supervision, target distillation, and relational distillation. Experiments on our newly collected unpaired MRI-fundus-biomarker dataset show that CGMD consistently improves fundus-based HTN prediction over standard distillation and non-graph imputation baselines, with ablations confirming the importance of clinically grounded graph connectivity. Code is available at https://github.com/DillanImans/CGMD-unpaired-distillation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_21809 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Clinical Graph-Mediated Distillation for Unpaired MRI-to-CFI Hypertension Prediction Imans, Dillan Bui, Phuoc-Nguyen Le, Duc-Tai Choo, Hyunseung Computer Vision and Pattern Recognition Retinal fundus imaging enables low-cost and scalable hypertension (HTN) screening, but HTN-related retinal cues are subtle, yielding high-variance predictions. Brain MRI provides stronger vascular and small-vessel-disease markers of HTN, yet it is expensive and rarely acquired alongside fundus images, resulting in modality-siloed datasets with disjoint MRI and fundus cohorts. We study this unpaired MRI-fundus regime and introduce Clinical Graph-Mediated Distillation (CGMD), a framework that transfers MRI-derived HTN knowledge to a fundus model without paired multimodal data. CGMD leverages shared structured biomarkers as a bridge by constructing a clinical similarity kNN graph spanning both cohorts. We train an MRI teacher, propagate its representations over the graph, and impute brain-informed representation targets for fundus patients. A fundus student is then trained with a joint objective combining HTN supervision, target distillation, and relational distillation. Experiments on our newly collected unpaired MRI-fundus-biomarker dataset show that CGMD consistently improves fundus-based HTN prediction over standard distillation and non-graph imputation baselines, with ablations confirming the importance of clinically grounded graph connectivity. Code is available at https://github.com/DillanImans/CGMD-unpaired-distillation. |
| title | Clinical Graph-Mediated Distillation for Unpaired MRI-to-CFI Hypertension Prediction |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2603.21809 |