Questions Every Sponsor Should Ask Their Imaging Partner

Introduction

Imaging is often discussed in terms of technology, biomarkers or endpoints. Those capabilities matter, but they are only part of the picture. The sponsor's real questions are whether the imaging strategy will generate evidence that supports confident decision-making, whether the workflow can deliver reliably, and whether the resulting insights will be available when they are needed most.

The following Q&A brings together key questions prompted by discussions at IPF Summit, spanning scientific evidence, operational delivery, treatment response and the realities of site delivery.

Ultimately, these questions are not about imaging itself. They are about generating more precise evidence and making better-informed clinical-development decisions.

1. Why is having the best imaging AI not enough?

A model can be scientifically strong and still fail to create value for a clinical trial if the surrounding workflow does not perform. Scans may arrive late, fail quality checks, be acquired inconsistently or miss the point at which an eligibility or study decision needs to be made. Sponsors should therefore assess both the quality of the analytics and the partner’s ability to operationalise them reliably. The practical question is not only “How well does the model perform?” but also “Can the resulting insight be delivered consistently, at scale and in time to support the trial?”

2. Where could handoffs put your trial at risk?

An imaging workflow may span site qualification, acquisition, transfer, quality control, central review, analytics and endpoint delivery. Every boundary is another interface where formats, timelines, quality requirements and responsibilities need to remain aligned. A multi-vendor model can work, but sponsors should know where the handoffs sit, what changes hands and how issues are escalated. The key question is who retains accountability for the complete imaging journey when a trial decision has a deadline, and how proactive they will be in putting things right.

3. Who owns the problem when imaging goes wrong?

Clear ownership matters when an eligibility read, image-quality query or endpoint delivery affects a live study timeline. Sponsors should test how a prospective partner detects problems, who coordinates resolution, how sites are supported and whether responsibility becomes fragmented between technology, operational and analytic providers. An integrated model can reduce boundaries, but good governance, clear service levels and transparent escalation remain essential.

4. Are you enrolling the patients your trial needs?

Baseline imaging can provide objective information about disease burden and patient characteristics that may not be captured fully by conventional eligibility criteria. Quantitative analysis can help sponsors understand the population entering the study, explore prognostic characteristics and monitor whether the enrolled cohort is changing across sites or over time. Imaging should sit alongside the clinical and statistical strategy, helping teams judge whether the recruited population is suited to the question the trial is designed to answer.

5. Are you getting enough value from the scans you already collect?

Trials already acquire imaging for eligibility, safety, central review or endpoint assessment. Quantitative analysis may allow those scans to contribute additional information about disease burden, cohort composition, treatment response or changes within specific anatomical compartments. The opportunity is not to create more data for its own sake. It is to identify analyses that are scientifically justified, appropriately planned and capable of informing a real development decision.

6. Could a more sensitive endpoint improve trial efficiency?

An endpoint that detects meaningful change with greater sensitivity may help a trial distinguish treatment effects more clearly. In some settings, that could support more efficient study designs or earlier confidence in an answer. Potential effects on sample size, duration or decision thresholds depend on the endpoint, disease, population, study design, statistical assumptions and regulatory context. Sponsors should ask for evidence of validation, reproducibility, biological relevance and performance in populations relevant to the proposed trial.

7. How do you know a treatment is working, and why it may be working?

Quantitative imaging can measure structural change over time and provide detail beyond a single whole-lung measure. Compartment-level analysis may explore patterns involving fibrosis, ground-glass opacity, airways, pulmonary vasculature or lung volume, depending on the disease and the validated analysis available. Used alongside clinical and biological data, this can help teams investigate both whether a response is emerging and which mechanisms may be contributing. Interpretation should remain grounded in the totality of the trial evidence.

8. What does a site-centric imaging approach look like?

Sites work within the demands of an already complex clinical study. Imaging processes should make qualification, training, image transfer and query resolution clear and practical rather than introduce avoidable burden. Sponsors should ask how quickly sites can be qualified, how acquisition quality is supported, how issues are resolved during enrolment and whether the imaging workflow is designed around the realities of site delivery.

Conclusion

The most important imaging-partner questions are not only about algorithms or platforms. They are about whether the scientific approach is appropriate, whether the workflow can support it and whether the resulting evidence will be ready when a trial decision needs to be made.

When used well, imaging can help sponsors understand who is entering a study, how disease is changing, whether a treatment response is emerging and where operational risk may sit. The objective is more precise evidence, better decisions and greater confidence across clinical development.