avatar

Julio Silva-Rodríguez

Trustworthy Machine Learning
Computer Vision · Medical Imaging
ETH Zürich
jusilva (at) ethz.ch


About

My research focuses on developing novel machine learning methods to address key technical barriers to the trustworthy deployment of AI in high-stakes application domains, such as computer vision and medical imaging. My goals are to:

  • Improve reliability under data constraints.
  • Control and monitor model performance at deployment.
Currently, my main research interests involve:
  • Specialized vision-language models pre-training and adaptation.
  • Conformal prediction and related frameworks providing statistically grounded guarantees.

My work spans a broad range of medical imaging domains and modalities, including: histopathology, volumetric radiology, retinal fundus imaging, and real-time endoscopic video.

I'm always open to new collaborations — drop me an email!

News

  • [06/26] 🆕 One paper accepted at MICCAI'26! 🎉 .
  • [03/26] 🆕 Appointed as Area Chair at MICCAI.
  • [01/26] 🆕 I have been awarded with the ETH Focus Grant! 💰
  • [10/25] I am starting a postdoc position at ETH Zurich 🛫 .
  • [07/25] I will participate in two MICCAI'25 tutorials: UQinMIA (conformal adaptation of medical VLMs) and FMLLM (foundation models for medical segmentation).
  • [06/25] We will be presenting 3 papers at MICCAI'25.
  • [05/25] One paper early-accepted at MICCAI'25 - see you in Daejeon! 🎉
  • [02/25] Our work on conformal prediction for VLMs has been accepted at CVPR'25.
  • [02/25] We got two papers accepted at IPMI'25 on medical VLMs! 🎉 .
  • [01/25] FLAIR, the first VLM for fundus images has finally been published at MedIA 📜 .
  • [05/24] Two papers accepted at MICCAI'24! 🎉 .
  • [03/24] Our work on few-shot CLIP adaptation is accepted to CVPR'24.
  • [08/23] Best paper award at the 1st MICCAI Workshop on foundation models! 🏅
  • [04/23] Awarded the Postdoctoral Merit Scholarship for Foreign Students (PBEEE) from the Fonds de recherche du Québec (FRQ)! 💰
  • [01/23] I started working as a Postdoctoral Fellow at ÉTS Montréal 🛫 .
  • [10/22] I defended my PhD Thesis in not-so-supervised learning for medical imaging 🎓 .
  • [06/22] Our paper about unsupervised brain lesion segmentation is accepted to MedIA 📜 .
  • [06/21] Our paper about weakly supervised WSI grading is accepted to JBHI 📜 .
  • [06/20] Our article on prostate histology grading - SICAPv2 dataset - is accepted at CMPB.
  • [09/19] Awarded with a 4-years PhD grant (FPI) from the Spanish Research Agency 💰 .
  • [09/19] Started my PhD studies under supervision of Prof. Valery Naranjo.

Selected Publications

  1. CVPR'25
    Conformal Prediction for Zero-Shot Models
    Julio Silva-Rodríguez, Ismail Ben Ayed, Jose Dolz
    CVPR, 2025.  
    PDF Code Project Page BibTex Poster
  2. MICCAI'26
    Trustworthy Endoscopic Super-Resolution.
    Julio Silva-Rodríguez, Ender Konukoglu
    MICCAI, 2026.  
    PDF Code
  3. CVPR'24
    A Closer Look at the Few-Shot Adaptation of Large Vision-Language Models
    Julio Silva-Rodríguez et al.
    CVPR, 2024.  
    PDF Code Project Page BibTex Poster Slides
  4. MedIA
    A Foundation Language-Image Model of the Retina (FLAIR): Encoding Expert Knowledge in Text Supervision
    Julio Silva-Rodríguez et al.
    Medical Image Analysis, 2025.  
    PDF Code Project Page BibTex Poster Slides -Open weights (>500 downloads/month)
  5. IPMI'25
    Full Conformal Adaptation of Medical Vision-Language Models
    Julio Silva-Rodríguez et al.
    IPMI, 2025.  
    PDF Code BibTex Slides -Oral
  6. MICCAI'25
    Trustworthy Few-Shot Transfer of Medical VLMs through Split Conformal Prediction.
    Julio Silva-Rodríguez, Ismail Ben Ayed, Jose Dolz
    MICCAI, 2025.  
    PDF Code
  7. IPMI'25
    A Reality Check of Vision-Language Pre-training in Radiology: Have We Progressed Using Text?
    Julio Silva-Rodríguez, Jose Dolz, Ismail Ben Ayed
    IPMI, 2025.  
    PDF Code Project Page BibTex Poster Slides
  8. MedIA
    Towards Foundation Models and Few-Shot Parameter-Efficient Fine-Tuning for Volumetric Organ Segmentation
    Julio Silva-Rodríguez, Jose Dolz, Ismail Ben Ayed
    Medical Image Analysis, 2025.  
    PDF Code Project Page BibTex Poster Slides
  9. CMPB
    Going deeper through the Gleason scoring scale: An automatic end-to-end system for histology prostate grading and cribriform pattern detection
    Julio Silva-Rodríguez et al.
    Computer Methods and Programs in Biomedicine, 2020.  
    PDF Code BibTex Dataset -SICAPv2 (+1.8K downloads)

Talks, Tutorials

  1. Uncertainty Quantification for Medical Foundation Models
    MICCAI'25 Tutorials: Uncertainty Quantification in Medical Image Analysis (UQinMIA)
    [09/25] Daejeon, Korea 🇰🇷 -
    Slides Code
  2. Foundational models for segmentation
    MICCAI'25 Tutorials: From Foundational to Multimodal Models for Medical Imaging (FMLLM)
    [09/25] Daejeon, Korea 🇰🇷 -
    Slides
  3. Towards Multi-Modal Foundation Models for Retinal Image Analysis
    APTOS-JSAIO'25 - invited faculty symposium on foundation models.
    [06/25] Osaka, Japan 🇯🇵 -
    Slides
  4. Full Conformal Adaptation of Medical Vision-Language Models
    IPMI'25 Oral Presentation
    [05/25] Kos, Greece 🇬🇷 -
    Slides
  5. Towards Trustworthy Vision-Language Models trough Conformal Prediction
    CentraleSupélec - Mathematics and Bioinformathics (MICS) research group.
    [01/25] Paris, France 🇫🇷 -
    Slides
  6. Conformal Prediction for Zero-Shot Models
    Sorbonne Université - Institut des Systèmes Intelligents et Robotique (ISIR).
    [01/25] Paris, France 🇫🇷 -
    Slides
  7. Foundation Models for Volumetric Medical Image Segmentation
    MICCAI'24 Tutorials: Foundation Models For Medical Imaging (FOMMIA)
    [10/24] Marrakech, Morocco 🇲🇦 -
    Slides
  8. Few-Shot Adaptation of Medical Vision-Language Models
    MICCAI'24 Spotlight Presentation
    [10/24] Marrakech, Morocco 🇲🇦 -
    Slides
  9. Hands-on: Few-Shot Adaptation of Medical VLMs
    Summer School on Deep Learning for Medical Imaging (DLMI) - ÉTS Montréal (MICCAI endorsed).
    [07/24] Montréal, Canada 🇨🇦 -
    Slides Code
  10. A Reality Check of Vision-Language Pre-training in Radiology
    Montréal 2024 Medical Imaging Workshop.
    [05/24] Montréal, Canada 🇨🇦 -
    Slides
  11. A Closer Look at the Few-Shot Adaptation of Large Vision-Language Models
    Interview at CVPR'24 Pre-show - Voxel51.
    [04/24] Online 🌐 -
    Video Slides Posts Medium
  12. FLAIR: A Foundation Vision-Language Model of the Retina
    Montréal 2023 Medical Imaging Workshop.
    [10/23] Montréal, Canada 🇨🇦 -
    Slides
  13. Towards Foundation Models and Few-Shot Parameter-Efficient Fine-Tuning for Volumetric Organ Segmentation
    MICCAIw'23: Int. Workshop on Foundation Models for General Medical AI (MedAGI)
    [10/23] Vancouver, Canada 🇨🇦 -
    Slides

Last updated: September 2026