Skip to main content
Clinical & Translational Science Institute
menu
Clinical & Translational Science Institute / Services and Support / AI ML Consultation Service

AI/ML Consultation Service

Artificial intelligence and machine learning (AI/ML) can turn large, complex biomedical data into knowledge faster and at a greater scale than traditional methods. But most health researchers are not familiar with these techniques, the field is shifting quickly, and applying AI/ML to health data carries risks, including opaque models, hidden bias, and data privacy concerns.

The CTSI AI/ML Consultation Service helps you decide whether AI/ML is the right tool for your question, choose trustworthy and defensible methods, find the data science resources you need, build the right team, and plan how your results will be used.

Jump to:
Return to:

Features

We work with investigators at any point in a project, including a first idea, a draft proposal, a resubmission addressing reviewer critiques, an analysis already underway, or a model you are trying to move into practice. Consultations are tailored to where you are in your project.

Idea development

  • Determine whether AI/ML is the right approach, or whether a conventional statistical or epidemiologic approach will serve you better
  • Frame a prediction, classification, phenotyping, or natural language processing task in terms that your data can support
  • Assess the feasibility based on the dataset you plan to use
  • Understand what infrastructure, software, and expertise a given approach will require

Writing or revising a proposal

  • Review analytic plans before submission, with attention to what reviewers scrutinize (model selection, training and validation strategy, sample size, bias)
  • Respond to reviewer critiques on a resubmission
  • Budget realistically for data acquisition, secure computing, software, and personnel
  • Build data management, sharing, and reproducibility plans that meet NIH DMSP requirements

Building a team

  • Connect with University of Rochester investigators, biostatisticians, informaticians, data scientists, and clinical domain experts working in your area
  • Identify co-investigators with demonstrated AI/ML expertise for grant applications
  • Learn what roles a project like yours typically needs
  • Link with trainee and student capstone projects through CTSI education programs and the Goergen Institute for Data Science and Artificial Intelligence

Analysis and interpretation

  • Think through model development, validation, performance metrics, calibration, and interpretability
  • Detect and mitigate bias in data, algorithms, and analytic methods
  • Match your data’s sensitivity to the right secure computing environment and toolset
  • Get a second opinion on results before they go into a manuscript

Implementation and translation

  • Apply implementation frameworks early so that a model is usable in the setting it is meant to serve
  • Plan for clinical decision support design, workflow fit, alert burden, and post-deployment monitoring
  • Anticipate ethical, regulatory, and equity considerations, including community values and trustworthiness in conjunction with Digital Innovation and Research Ethics Consultation Service
  • Connect with University of Rochester enterprise-level informatics and the UR Health Lab when a tool is developed for the clinical environment

Get Started

Know what you need? Complete the AI/ML Consultation Service request form. A member of our team will respond within one week to schedule a consultation.

Not sure yet? Email the CTSI Research Help Desk at CTSI Research Help Desk, and we will help you identify the services that fit your project. Consultations are available at no cost to University of Rochester faculty, staff, trainees, and students.

Timing

Reach out as early as you can. For grant submissions, contacting us three or more months before your deadline gives you time to refine the analytic plan, secure data access, and bring collaborators onto the application. We are glad to help as the deadline approaches, though the scope of what we can do narrows.

What to Prepare

  • A short description of your research question
    • Whatever you have already written
    • Specific aims
    • Draft proposal
  • Summary statement
  • Data you have or hope to use
  • List of current research team members

Nothing needs to be polished; early and rough is easier to help with than late and fixed.

Resources

AI/ML Research Methods Forum (launching soon)

A multidisciplinary forum where investigators present new or developing research with AI/ML components and receive structured input from methods, informatics, ethics, and clinical experts. Modeled on the CTSI Research Methods Forum. All investigators who submit a consultation request are invited to attend.

Analytics Colloquia

Twice-yearly presentations of translational projects to the research community, illustrating new methods and creating openings for collaboration. At least one event each year focuses on AI methods for translational research.

Data, Computing, and Relevant Support Services

  • Secure Environment for Research Data Analytics (SERDA) — A secure enclave for analysis of limited, identified, and restricted-access datasets, including secure cloud resources for AI/ML high-performance computing.
  • EHR data — A process for obtaining patient data from the EHR, including self-service access to a "rapid research ready" deidentified dataset and OMOP-mapped clinical data for cohort discovery, feasibility assessment, and model development.
  • CIDUR (Clinical Imaging Data for UR Researchers) — A web-based platform for managing clinical imaging data (MRI, CT, X-ray, ultrasound) in IRB-approved studies, with secure storage, access controls, de-identification, quality assessment, and a DICOM viewer.
  • Enterprise informatics tools — Institutionally supported, HIPAA-compliant large language model environments (e.g., Qualified Health) and emerging clinical AI tools, in coordination with other URochester Medicine technical teams.
  • Informatics Consultation - The CTSI Informatics team can help you plan, access, analyze, and manage data throughout the course of your studies. Whether you’re working with clinical, population health, or basic biomedical data, this team has the tools and expertise to get you what you need.
  • Biostatistics, Epidemiology, and Research Design (BERD) - The BERD group is a central source of support for translational investigators in study design, data analysis, and grantsmanship, and fosters the development of novel statistical and analytic methods for clinical and translational science research.
  • Research IT Support Service - The Office of Research IT provides basic desktop support to all faculty, staff, and students at the University of Rochester School of Medicine and Dentistry, as well as extended technical support to several groups and departments.
  • Research Help Desk – CTSI’s central contact for all research-related questions. Connects researchers to appropriate service providers and resources.

Education and training

Graduate courses in biomedical informatics — the clinical informatics elective for postgraduate trainees, the Certificate of Advanced Study in Biomedical Data Science, and periodic hack-a-thons pairing clinicians with data scientists.

Contact

To ask questions and learn how the AI/ML Consultation Service can accelerate your research, contact the Research Help Desk.

Leaders

Erika Ramsdale, M.D., M.S.
Strategic Director, CTSI Informatics
Erika_Ramsdale@urmc.rochester.edu

Adam Dziorny, M.D., Ph.D., FAMIA
Associate Strategic Director, CTSI Informatics
Adam_Dziorny@urmc.rochester.edu

Jack Chang, Ph.D.
Operations Director, CTSI Informatics
Jack1_Chang@urmc.rochester.edu