Join the Deep Learning for Precision Health Lab at UT Southwestern to build next-generation AI for medicine using Bayesian and probabilistic modeling, with direct access to large, deeply phenotyped datasets and clinical partners across UT Southwestern Medical Center, Children’s Medical Center Dallas, Parkland Hospital, and the O’Donnell Brain Institute. These roles are ideal for researchers who have recently (or will soon) completed a PhD (typically ≤2 years from degree; ABD considered). Based in Dallas—one of the largest, most vibrant, and fastest-growing cities in the U.S.—fellows work closely with Prof. Albert A. Montillo, PhD (Associate Professor, tenured, Fellow of MICCAI / IEEE / SPIE / ASNR / ISMRM / OHBM) and collaborate with statisticians, neurologists, radiologists, psychiatrists, and neuroscientists on clinically grounded problems—aimed at high-impact publications and deployable, uncertainty-aware methods.
________________________________________
Project tracks (pick one or blend across):
1. Bayesian Deep Multimodal Fusion & GNNs:
Integrate multi-contrast MRI & PET with electrophysiology, EHR/clinical data, and omics (genomics/proteomics) using deep Bayesian fusion models and graph neural networks. Emphasize uncertainty quantification in predicting disease progression and treatment response (Applications in Parkinson’s, AD, ASD, epilepsy, depression).
2. Foundation Models for Medical Imaging:
Pretrain and fine-tune foundation models on large-scale image datasets (10k–100k+ subjects). Combine data-driven and prior-informed modeling for explainable transfer learning with principled handling of uncertainty across sites.
3. Bayesian Causal Discovery & Probabilistic Graphical Models:
Develop new Bayesian methods for inferring causal brain connectivity from neuroimaging and interventional data, leveraging prior knowledge, uncertainty estimates, and model evidence in developmental disorders (e.g., ASD, epilepsy).
4. Bayesian Reinforcement Learning for Neuromodulation:
Combine computational neuroscience models with data-driven foundation models using Bayesian Model Averaging and probabilistic policy optimization to guide closed-loop neuromodulation under uncertainty.
5. Multimodal Speech + Imaging for Early Dementia:
Build foundation models over audio, linguistics, and imaging data for early dementia diagnosis, emphasizing probabilistic representations and uncertainty calibration to support clinical deployment.
________________________________________
How your career will grow:
• Design novel methods (FMs, deep Bayesian fusion, GNNs, probabilistic causal discovery, RL) and run rigorous experiments at scale.
• Lead manuscripts to AISTATS, NeurIPS, ICLR, ICML, MICCAI, IPMI, TMI, MedIA, TPAMI, CVPR.
• Collaborate with clinical teams for problem definition, data access, and validation.
• Contribute to open-source codebases, reproducible pipelines, and trustworthy AI practices.
________________________________________
Required Qualifications:
1. PhD (or near completion) in CS, ECE, Applied Math, Computational Physics, BME, Bioinformatics, Statistics, or a related field, with experience in machine learning and analysis of imaging, omics, audio, text, or clinical data.
2. strong background in probability, statistics, and mathematical modeling.
3. Proficient in coding mathematical models in Python and experience with deep learning coding (PyTorch/TensorFlow)
4. Major contributions in top-tier peer-reviewed venues (AISTATS, NeurIPS, ICLR, ICML, MICCAI, IPMI, CVPR, TPAMI, TMI, MedIA, Nature Comms, etc.).
5. Enthusiasm to drive clinical research projects to completion with clear, reproducible results.
________________________________________
Preferred (nice to have, one or more of these):
Experience with Bayesian methods, graphical models, or variational inference.
Foundation model development; deep multimodal fusion; speech analysis / computational linguistics; explainable AI;
Causal analysis; computational neuroscience; genomics/proteomics analysis; C++/CUDA, MATLAB experience; prior clinical collaboration.
________________________________________
Appointment and Support:
Full-time position based in Dallas, TX, with competitive salary and full UTSW benefits. Initial appointment for 1 year, renewable based on performance. Fellows should plan for a minimum 2-year commitment to support impactful publications and career growth. U.S. citizens are encouraged to apply; visa sponsorship available for exceptional international candidates. Target start date: 2025; later start dates considered.
________________________________________
For Consideration:
Email Albert.Montillo@UTSouthwestern.edu with subject “Postdoc-Applicant-ISBA” and include:
1. CV
2. Contact information for 3 references
3. Up to 3 representative publications
4. Your preferred track(s) and start window
Positions are open until filled; review begins immediately.
URL: https://montillolab.org/#PostdoctoralFellowshipPositions
