Scientific & Data Advisory for Medical Imaging R&D
I advise medical imaging AI teams on complex R&D questions, bringing scientific rigor to decisions around data, experimentation, and evaluation.
With a PhD in Computer Vision and over ten years of experience in applied medical imaging R&D, I combine actionable literature review with deep data understanding to help teams frame the right questions, design meaningful experiments, define robust evaluation strategies, and draw conclusions supported by the available evidence.
From R&D Questions to Evidence
A structured overview of how I support medical imaging AI teams.
Collaboration Format
I work with medical imaging AI teams through targeted engagements or ongoing advisory, depending on the question and level of support needed.
- Targeted engagements address a specific R&D question or decision, from literature and data assessment to experimental design and evaluation.
- Ongoing advisory provides regular scientific and data guidance as R&D questions evolve, helping teams challenge assumptions, review evidence, and make well-founded decisions over time.
Technical Framing & Feasibility
Clarifying scientific questions, objectives, assumptions, and constraints to support sound R&D decisions.
Data Understanding & Qualification
Assessing what the available data can reliably support, and identifying limitations, biases, and sources of variability that may affect development and evaluation.
Experiment & Evaluation Strategy
Designing experiments & evaluation strategies so that results are interpretable, comparable, and aligned with the intended use.
Literature Review & Analysis
Turning the state of the art into actionable R&D insight: established evidence, uncertainties, and findings relevant to the question at hand.
Evidence & Decision Documentation
Documenting assumptions, evidence, and technical decisions so that the rationale remains traceable and defensible over time.
Case Study
An in-depth analysis of a public dataset of 112,120 images across 30,805 patients surfaced a risk that is easy to miss and costly to ignore: a single patient contributing up to 184 images, creating a significant data-leakage risk when splits are not handled at the patient level.
This is one of several findings that directly inform how the dataset should be used and how evaluation should be designed. The full analysis is available on the case study page.
Who I Work With
- MedTech startups and scaleups developing AI-based products
- Medical imaging and healthcare teams advancing AI R&D
- R&D leaders and teams seeking scientific and data guidance on development and evaluation questions
What This Enables
- Evidence-based R&D decisions
- Stronger experimental and evaluation rigor
- Clearer understanding of data qualities and limitations
- More traceable technical reasoning
Why Work with Me
- Deep Technical Authority
PhD-level expertise in computer vision, combined with over ten years of applied medical imaging R&D. - Rigor for Defensible Decisions
A structured approach to assessing evidence, challenging assumptions, and grounding technical choices in rigorous scientific reasoning. - Experience at the R&D-Reality Interface
Hands-on involvement in projects where scientific questions, data realities, and real-world clinical requirements must align. - Decision-Oriented Communication
Complex technical issues translated into structured, decision-ready insights for technical and non-technical stakeholders alike.
Facing a complex R&D question?
Let’s examine the evidence, data, and evaluation strategy behind your next decision.
