Júlio G. Domingues

Radiologist · Computer scientist

Júlio Guerra Domingues

I work at the intersection of radiology and machine learning, building and validating deep-learning models for medical imaging.

Assistant Professor, Federal University of Minas Gerais (UFMG)
Visiting Research Fellow, Boston University

Publications Get in touch

Júlio Guerra Domingues

About

I am a board-certified radiologist and a computer scientist. My clinical work is in neuroradiology and musculoskeletal imaging, and my research applies deep learning and computer vision to imaging problems — from disease detection and classification to the clinical validation of AI models.

I teach medical students and radiology residents as an Assistant Professor at the UFMG School of Medicine, where I am also a PhD candidate in Public Health, studying deep-learning models for knee osteoarthritis. I am currently a Visiting Research Fellow at Boston University. I hold an MD and a BSc in Computer Science from UFMG and an MSc in Applied Adult Health Sciences.

Research

My focus is translating imaging-AI research into tools that hold up in clinical practice. Current and recurring threads:

  • Deep learning and computer vision for medical imaging
  • AI-assisted disease detection and classification
  • Clinical validation and translation of AI models
  • Population-based imaging studies (ELSA-Brasil MSK)
  • Neuroradiology — brain MRI and dementia
  • Musculoskeletal imaging — knee osteoarthritis

Selected publications

Peer-reviewed and indexed work. A complete list is on ORCID.

  1. Domingues JG, Veloso AA, Telles RW, Barreto SM. A clinical-epidemiological tool for identifying prevalent radiographic knee osteoarthritis: data from ELSA-Brasil MSK. Osteoarthritis and Cartilage. 2026;34(6):947–948. doi:10.1016/j.joca.2026.03.019
  2. Carvalhido L, Domingues J, Miranda L, et al. 3D Convolutional Neural Networks for Fazekas Scale Classification in Brain MRI. Anais do XVII Congresso Brasileiro de Inteligência Computacional. 2025:1–7. doi:10.21528/CBIC2025-1191915
  3. Domingues JG, Costa-Silva L, Machado AM, et al. External validation of a pre-trained convolutional neural network for osteoarthritis diagnosis in knee radiographs of ELSA-Brasil MSK. Osteoarthritis and Cartilage. 2024;32:S70–S71. doi:10.1016/j.joca.2024.02.103
  4. Domingues JG, Araujo DC, Costa-Silva L, et al. Development of a convolutional neural network for diagnosing osteoarthritis, trained with knee radiographs from ELSA-Brasil Musculoskeletal. Radiologia Brasileira. 2023;56(5):248–254. doi:10.1590/0100-3984.2023.0020-en

Additional work appears in the proceedings of the European Congress of Radiology (ECR 2018–2020).

Software & projects

UNOTE

Critical-findings notification system for radiology. Registered software, 2023.

home.unote.com.br →

RADsmart

Evidence-based radiology tools and standardized report lexicons. A side project, in development.

About RADsmart →

Education & training

  • 2024 – Present PhD Candidate in Public HealthFederal University of Minas Gerais (UFMG)
  • Jan 2026 – Present Visiting Research FellowBoston University
  • 2025 BSc in Computer ScienceFederal University of Minas Gerais (UFMG)
  • 2023 MSc in Applied Adult Health SciencesFederal University of Minas Gerais (UFMG)
  • 2020 Board Certification in Radiology & Diagnostic ImagingBrazilian College of Radiology (CBR)
  • 2019 Medical Residency in Radiology & Diagnostic ImagingMadre Teresa Hospital
  • 2015 Doctor of Medicine (MD)Federal University of Minas Gerais (UFMG)

Additional training in AI

  • 2024 CS50’s Introduction to Artificial Intelligence with PythonHarvard University (120h, online)
  • 2020 Deep Learning SpecializationDeepLearning.AI (72h, online)
  • 2020 AI for Medicine SpecializationDeepLearning.AI (80h, online)
  • 2017 – 2019 European Course of Diagnostic & Interventional Neuroradiology (ECNR-LATAM)European Society of Neuroradiology (152h)

Awards & honors

  • 2022 1st Place — Best Paper (Magna Cum Laude)51st Brazilian Congress of Radiology (CBR)
  • 2022 1st Place — Best Paper in AI, Innovation & Teleradiology51st Brazilian Congress of Radiology (CBR)
  • 2018 Medal of Honor — Radiology ResidentRadiology Society of Minas Gerais (SRMG)
  • 2018 Invest in the Youth AwardEuropean Society of Radiology (ESR)

Contact

For research, teaching, or collaboration, the best way to reach me is by email.