How One Mathematician’s PhD Insight Built a Global Imaging CRO – and Changed How Drug Trials Measure What Matters

In a cluttered computer science lab at the University of Leeds sometime in the early 2000s, a mathematician from Kazakhstan was staring at MRI scans and asking a question that most radiologists at the time had given up trying to answer: exactly how much inflammation was visible in that joint – not qualitatively, not impressionistically, but as a number that a drug company could take to the FDA?

Dr. Olga Kubassova was pursuing a PhD in Computer Science, building novel algorithms to quantify inflammation in medical images. She was not a radiologist. She was a mathematician, and that distinction mattered enormously. Where clinicians saw shading and texture, she saw signals to be extracted, normalized, and turned into data points. Where others saw the problem of reading MRI scans as fundamentally subjective, she saw an engineering challenge waiting for a sufficiently precise solution.

Her PhD work earned her the Sir Peter Thompson Enterprise Award — she was its first winner — along with £5,000 in seed funding. It was a small sum. The idea it seeded was not small at all.

By 2007, the pharmaceutical industry had developed a significant problem it had not yet fully named. With success of Humira® and other autoimmunity focused blockbuster drugs, trials proliferated. As regulatory agencies demanded more rigorous, reproducible imaging data to assess patient respond and drug impact early, the gap between what a general contract research organization (CRO) could provide and what a trial actually needed – in terms of specialized imaging interpretation, platform infrastructure, and compliance oversight – had grown wide enough to be dangerous.

Sponsors were losing trials and opportunities to develop competitive drug candidates not because their drugs failed, but because the imaging data was unreliable and did not give strategic go/no-go decisions early. Sites used different scanners. Readers applied criteria inconsistently. There was no systematic imaging charter development process. Nobody was performing site qualification imaging checks before data started flowing. And when problems emerged mid-trial, nobody had a playbook for imaging trial rescue – no specialist imaging CRO to call, no established methodology for salvaging a compromised imaging dataset.

Back then, academic labs and radiology groups started writing various guidance documents and pulling image storage solutions to assist. The absence of a credible clinical trial imaging vendor with reliable cloud-based platform for management of images in real-time meant that imaging endpoints, which should have been the most objective evidence in a trial, were often its weakest link. Sponsors who had chosen the cheapest or most convenient imaging vendor were discovering that clinical trial imaging outsourcing to a general-purpose provider created more problems than it solved. What the market needed was not more outsourcing. It needed specialized outsourcing and smart platform for management of imaging data – to a purpose-built imaging core lab with deep scientific credentials and a systematic approach to every step of the imaging workflow.

Kubassova had spent her academic years solving the technical side of exactly this problem. In 2007, she founded Image Analysis Group – IAG – and set out to solve the operational and regulatory side too.

The founding thesis was specific: the pharmaceutical industry needed a dedicated imaging CRO. Not a general CRO with an imaging department. Not a radiology practice moonlighting in clinical trials. A company whose entire reason for existence was the design, management, and analysis of imaging endpoints in drug development studies. An imaging core lab that could serve as the institutional backbone of a clinical trial’s most sensitive measurement system.

Success of such operational fully relies on ability to build enterprise level platform for management of imaging data at global scale. IAG embarked on building DYNAMIKA™, the first and now industry leading enterprise level platform which became Operating System for IAG and IAG’s partners.

The first customer was a large pharmaceutical company developing drugs for rheumatoid arthritis — exactly the therapeutic territory where Kubassova’s PhD algorithms had their roots. From the beginning, IAG’s imaging CRO services combined proprietary technology DYNAMIKA™ with deep clinical expertise, a model that would define the company for the next two decades.

To understand why a specialized imaging CRO exists, it helps to understand what imaging actually does in a clinical trial and how much can go wrong when it’s handled poorly.

In an oncology imaging clinical trial, imaging endpoints are often the primary evidence that a drug is working. Under response criteria like RECIST 1.1, a radiologist must measure the longest diameter of target lesions across multiple time points and determine whether a patient’s tumor has responded, progressed, or remained stable.

That assessment – response, stable disease, or progression – can determine whether a trial succeeds or fails, whether a drug gets approved, and whether patients get access to a new treatment.

But the measurement is not automatic. It requires a trained reader, a standardized protocol, consistent scanning parameters, and a mechanism to prevent the reader from knowing whether the patient is receiving the drug or a placebo. That last requirement – blinding – is why blinded independent central review has become a regulatory expectation rather than a nice-to-have. The FDA and EMA have made clear in multiple guidance documents that for trials where imaging is the primary efficacy endpoint, blinded independent central review is expected. For accelerated approval pathways and certain oncology indications, BICR is essentially mandatory.

The imaging core lab sits at the center of this system. An imaging core lab receives images from trial sites around the world, manages their quality, assigns them to blinded readers, and aggregates the results. A well-run imaging core lab is the difference between an imaging dataset that holds up to regulatory scrutiny and one that gets questioned during an advisory committee meeting.

This is the problem Image Analysis Group was built to solve – and has spent nearly two decades solving, one trial at a time.

The tool that became the operational center of IAG’s imaging core lab services was DYNAMIKA™ – a proprietary platform for centralized image management, stakeholder workflows, reader workflow, endpoint adjudication, and compliance audit trails. DYNAMIKA™ is the technological expression of IAG’s imaging core lab services: a system designed from first principles to support the full lifecycle of clinical trial imaging outsourcing, from protocol design through regulatory submission.

DYNAMIKA™ was not built in a single version. It evolved through seven major iterations over approximately eighteen years, each one responding to what the company learned from running actual trials.

The early versions were essentially sophisticated in-house tools: algorithms that could quantify MRI data for inflammatory and musculoskeletal indications with a reproducibility that manual reading could not match. As IAG moved from serving single sponsors to running multi-site global trials, the platform had to scale accordingly.

By 2011, IAG made a move that was genuinely ahead of its time in clinical imaging: DYNAMIKA™ went fully to the cloud. This was not a trivial decision. Clinical imaging data carries privacy obligations – HIPAA, GDPR, national data residency laws – that made many sponsors nervous about cloud infrastructure. IAG solved the compliance architecture problem and moved anyway, gaining years of operational experience with cloud-based centralized image management at a time when competitors were still managing local servers.

The cloud migration also resolved a practical problem that any global imaging trial faces: how do you get enormous DICOM image files from a site in Tokyo or Nairobi to a Reader expert who might be based in London or Boston in a timeframe that doesn’t delay trial timelines? The answer, in 2026, is obvious. In 2011, it required genuine technical and regulatory innovation.

The seventh generation of DYNAMIKA™ – DYNAMIKA V.7 – launched in January 2025 with SOC II Type I compliance and BSI surveillance audit certification. The platform’s innovation track record was recognized publicly when it was named an Award Category Winner for Innovation at the 2024 Pharmaceutical Technology Excellence Awards.

DYNAMIKA V.7 supports the full spectrum of imaging core lab services, from ingestion and quality control through AI-assisted reading, assessment, and regulatory-grade audit trails compliant with FDA 21 CFR Part 11. Trial critical elements such as Turn-Around-Time (TAT) for image QC and Image Review are fully addressed through the cloud data management and full connectivity between all stakeholders. IAG allows for as little as 12 hours Turn Around Time for image Quality Control. The statistics from phase III trials show that even with dual-read and adjudication, the 70% of the images can be read with Turn-Around-Time (TAT) less than 3 days, with 50% data being reviewed within 24 hours.

Speaking of 21 CFR Part 11: the regulatory landscape for clinical trial imaging has grown substantially more demanding since IAG was founded, and understanding that landscape is essential context for understanding why imaging CRO selection has become a strategic decision rather than a procurement exercise.

21 CFR Part 11 governs electronic records and electronic signatures in FDA-regulated studies. For imaging trials, this means every image, every reader annotation, every adjudication decision, and every change to a measurement must be captured in an audit trail that demonstrates who did what, when, and whether they had authority to do it. A platform that is not built around 21 CFR Part 11 compliance from the ground up is a platform that will create problems during an FDA inspection – potentially after the trial is already complete.

Alongside 21 CFR Part 11, the updated ICH E6(R3) Good Clinical Practice guidelines reinforce the importance of data integrity, risk-based monitoring, and qualified personnel throughout the imaging workflow. For imaging, this translates directly into requirements around site qualification imaging -verifying that a trial site’s scanner, protocol, and technical staff meet the trial’s requirements before the first patient is enrolled. Fail to qualify a site properly, and you may accumulate months of images that are non-evaluable by the time someone reviews them.

Imaging charter development is the document-level expression of all these requirements. A well-constructed imaging charter specifies the imaging modalities, acquisition protocols, reader qualifications, response criteria (whether RECIST 1.1, iRECIST, mRECIST, RANO, RANO 2.0, Lugano, RAMRIS, DEMRIQ or any of a dozen others), adjudication procedures, and data transfer requirements for the trial. It is the contract between the sponsor, the imaging CRO, and the sites and it needs to be written before a single patient is scanned.

IAG’s experience developing imaging charters across more than 700 global clinical trials across Phase I through Phase IV represents an institutional depth that cannot be replicated quickly. The company has written charters for trials in solid tumors using RECIST 1.1 and volumetric tumor assessment; for hematologic malignancies using Lugano and Deauville criteria; for neuro-oncology trials using RANO and RANO 2.0; for musculoskeletal trials using MOAKS, 11-point, DEMRIQ, RAMRIS and SPARCC; for liver disease trials using elastography and fat fraction quantification. The breadth itself is a form of expertise.

The vocabulary of clinical trial imaging can be bewildering to those who encounter it for the first time. RECIST 1.1. iRECIST. mRECIST. RANO. RANO 2.0. Lugano. Cheson. Deauville. RAMRIS. SPARCC. MOAKS. PI-RADS. BI-RADS. Each of these is a response criteria framework – a set of rules for turning images into decisions.

Understanding why so many frameworks exist illuminates why specialized imaging CRO services are necessary.

RECIST 1.1 (Response Evaluation Criteria in Solid Tumors) is the standard for most solid tumor oncology imaging clinical trials. It defines how to select target lesions, how to measure them, and what thresholds of change constitute response, stable disease, or progression. RECIST 1.1 is well-validated and widely accepted but it was not designed for immunotherapy, where tumors sometimes appear to grow before they shrink (pseudoprogression). iRECIST was developed specifically for immunotherapy trials, allowing for unconfirmed progression before calling a patient’s disease truly progressive. mRECIST, a modified version, applies to hepatocellular carcinoma, where viable tumor tissue rather than total lesion size is the relevant measurement.

The practical implication for an imaging core lab is that a single oncology imaging program may require readers trained in RECIST 1.1 for some patients, iRECIST for others receiving immunotherapy combinations, and mRECIST for hepatocellular carcinoma subsets all within the same trial. Managing this complexity – training readers on multiple criteria, maintaining blinding across criteria, adjudicating discordant reads is one of the reasons that oncology imaging clinical trial management has become one of the most technically demanding services an imaging CRO provides.

Volumetric tumor assessment is an emerging approach that measures three-dimensional tumor volume rather than the longest diameter used in RECIST 1.1. Evidence suggests volumetric tumor assessment may detect treatment effects earlier and more sensitively than linear measurement particularly for irregularly shaped lesions. AI-assisted tools are making volumetric tumor assessment increasingly practical at scale.

The RANO criteria apply to brain tumors: gliomas, glioblastoma, brain metastases. RANO 2.0, updated in recent years, addresses issues including pseudo progression and treatment-related changes that can mimic tumor growth on MRI. Neuro-oncology imaging requires not just expertise in RANO criteria but deep familiarity with the patterns that make brain tumor imaging particularly challenging to read and particularly consequential to get right.

Musculoskeletal trials have their own ecosystem: RAMRIS for rheumatoid arthritis MRI, PSAMRIS for psoriatic arthritis,  SPARCC for sacroiliac joint inflammation in ankylosing spondylitis, mSASSS for spinal structural damage, MOAKS for knee osteoarthritis. IAG’s origins in quantitative MRI for inflammatory arthritis gave the company a foundation in this space that remains central to its identity. The MRI endpoint in a rheumatoid arthritis trial is fundamentally different from a MRI endpoint in a glioblastoma trial — different anatomy, different pathophysiology, different measurement tools, different reader expertise — and an imaging core lab that treats them as interchangeable will produce inferior data in both.

An imaging core lab that can navigate all of these frameworks simultaneously and that can build reading panels of physicians who are expert in each — is not a commodity service. It is a specialized capability built over years of practice.

IAG began with musculoskeletal autoimmune disease. It has grown to cover virtually every therapeutic area where imaging plays a meaningful role in drug development.

In oncology imaging clinical trials, IAG works across solid tumors – breast, colorectal, prostate, lung, pancreatic, sarcomas, head and neck as well as hematologic malignancies including lymphoma and myeloma, and neuro-oncology. The company applies RECIST 1.1, iRECIST, mRECIST, volumetric tumor assessment, and PET imaging metrics including standardized uptake value (SUV) analysis across FDG PET/CT and PET/MRI studies.

PET imaging clinical trials present specific challenges. PET is inherently quantitative – standardized uptake values are the primary output but that quantification is only meaningful if acquisition protocols, reconstruction parameters, and image quality are tightly controlled. SUV-based tumor response assessment requires consistent patient preparation, injection timing, scanner calibration, and image reconstruction across all trial sites. An imaging core lab running a multi-site PET imaging clinical trial must ensure that a scan acquired in Shanghai is directly comparable to one acquired in Chicago. Without a rigorous imaging core lab service managing harmonization, SUV-based tumor response assessment data becomes incomparable across sites and the primary endpoint collapses.

For oncology imaging specifically, PET imaging enables whole-body metabolic tumor burden assessment that anatomical imaging alone cannot provide. FDG PET/CT tumor response assessment under Deauville or Cheson criteria in lymphoma trials, and tau PET in neurodegenerative programs, represent MRI endpoint work at the outer boundary of what imaging endpoints can measure requiring an imaging CRO with both the technical infrastructure and the expert reader network to execute these protocols reliably across global sites.

In rare disease clinical trial imaging, IAG works in genetic disorders, Duchenne muscular dystrophy (DMD), and other rare disease indications where quantitative MRI may be the only available measure of disease progression or treatment effect. The challenge in rare disease imaging is that validated response criteria may not exist – the imaging CRO must help design novel endpoints, validate the measurement approach, and build the regulatory case for accepting imaging as a primary outcome. This requires a level of scientific engagement that goes well beyond reading images.

For sponsors in rare disease programs, the imaging vendor is often a scientific collaborator more than a service provider. A rare disease imaging CRO selection is not primarily about throughput or platform cost – it is about finding a partner that can co-develop the MRI endpoint or quantitative MRI biomarker from first principles, validate it in a patient population that may number in the dozens rather than thousands, and present the measurement methodology to regulators with enough scientific rigor to support a breakthrough therapy or accelerated approval pathway. That combination of quantitative MRI expertise and regulatory fluency is rare, and it is not something a general clinical trial imaging vendor can offer.

In neurology, IAG works across traumatic brain injury, stroke, and psychiatric indications. In autoimmune disease, the company covers scleroderma, systemic lupus erythematosus, myositis, and CAR-T therapy monitoring. In metabolic liver disease – NASH/MASH, obesity-related hepatic steatosis – multiparametric MRI including MR elastography and fat fraction quantification has become the central tool for non-invasive efficacy assessment. The multiparametric MRI protocol for NASH trials is one of the more technically demanding imaging programs in drug development, requiring acquisition of multiple quantitative maps – T1, T2*, proton density fat fraction, elastography within a single breath-hold session, and analysis pipelines that extract liver fat and fibrosis metrics reproducibly across scanner manufacturers and field strengths.

For trials that rely on bone health analysis, such as trials in obesity, osteoporosis, and metabolic disorders, IAG provides end-to-end imaging CRO services with DXA solution that helps the sites to standardize their equipment and calibrate it during the trial. IAG support all FDA required DXA analyses and brings sponsors an opportunity to develop truly differentiated drug candidates, especially as GLP-1 gaining momentum.

Women’s health is represented through IAG’s partnership with Ferring Pharmaceuticals, announced in 2023, focused on AI-driven medical image analysis in reproductive health – a therapeutic area where AI imaging tools are creating new possibilities for objective, reproducible measurement in a space that has historically relied on subjective interpretation.

The conversation about AI in clinical trial imaging is often framed in terms of replacing radiologists. That framing misses the more immediate and more interesting story: AI imaging tools are making it possible to extract information from images that human readers simply cannot see.

Radiomics is the extraction of large numbers of quantitative features from medical images – texture, shape, intensity distribution, spatial relationships and the use of those features in predictive models. A radiomic signature derived from a baseline CT scan might predict which patients will respond to a given therapy, enabling trial enrichment. A radiomic feature extracted from serial MRI scans might detect treatment response weeks before conventional RECIST 1.1 assessment would register a change. Radiomics clinical trial applications are still emerging from the research phase into validated endpoint use, but they represent a significant near-term opportunity for sponsors willing to invest in the data infrastructure.

For radiomics to contribute to a regulatory submission, the imaging data underlying the radiomic features must be collected under a rigorous imaging protocol with consistent acquisition parameters – precisely the kind of centralized image management and site qualification imaging work that an imaging core lab performs on every trial. Radiomics and standardized data collection are inseparable; sponsors who try to conduct post-hoc radiomics analysis on imaging data collected without a radiomics-aware imaging charter development process typically find the features are confounded by scanner and protocol variability. The imaging CRO that integrates radiomics methodology at the protocol design stage – rather than retrospectively is the one that produces radiomics data worth analyzing.

Pathomics extends the same approach to histopathology images – whole-slide images from tissue biopsies analyzed with algorithms that can quantify inflammatory cell density, tumor microenvironment features, and tissue architecture at a scale no human pathologist could achieve manually.

IAG’s partnership with Takeda represents one of the most prominent examples of AI imaging applied in clinical development: AI-powered histopathological image analysis in ulcerative colitis, a therapeutic area where tissue-level inflammation quantification is central to demonstrating drug efficacy. The ability to measure microscopic inflammation objectively and at scale is the kind of capability that changes what is possible in a clinical trial.

In oncology, IAG’s partnership with Picture Health – announced in May 2025 – focuses on AI-driven imaging endpoints in oncology clinical trials, bringing together IAG’s infrastructure for centralized image management and regulatory-grade BICR with Picture Health’s AI tools for image analysis. The collaboration with Better Medicine similarly targets AI collaboration in oncology imaging.

These partnerships reflect a model of the imaging CRO as an integrator not just a vendor that brings together clinical imaging expertise, regulatory compliance infrastructure, and AI capabilities to address sponsor problems that no single capability could solve alone.

It is worth noting what AI imaging does not yet do. It does not replace the regulatory requirement for blinded independent central review on primary imaging endpoints. A sponsor using AI tools for tumor response assessment still needs the results reviewed by blinded independent readers and documented in a compliant audit trail. The AI accelerates and enriches the process; it does not circumvent the governance structure that makes the data trustworthy.

Seven hundred clinical trials in nineteen years. Phase I through Phase IV. Sponsors ranging from the world’s largest pharmaceutical companies to small biotechs conducting first-in-human studies.

IAG’s global footprint is operationally real, not just a dot-on-a-map exercise. Headquarters in London, UK. Offices in the US and India. Operations spanning China, Europe, Australia, and Africa. A global site network that can onboard imaging sites in virtually any country and provide 24/7 expert support to trial teams working across time zones.

The global dimension of clinical trial imaging outsourcing has grown in importance as trial operations themselves have globalized. Clinical trial imaging outsourcing is no longer simply a question of which imaging core lab has the best platform – it is a question of which imaging CRO has the operational reach, language capabilities, and regulatory knowledge to function effectively in Lagos, Seoul, Buenos Aires, and Warsaw simultaneously. A Phase III oncology trial in 2026 might enroll patients at 200 sites across 40 countries. Each of those sites needs to be qualified to the imaging charter’s standards, trained on the imaging protocol, and integrated with the central imaging platform. When a site in Malaysia has a question about their scanner protocol at 3 AM London time, the imaging vendor needs to have someone available who can answer it.

The compliance stack that underpins IAG’s global operations is similarly serious: FDA 21 CFR Part 11 for electronic records; ISO 13485 for quality management; GDPR and HIPAA for data privacy; SOC 2 Type II for cloud security; BSI surveillance audit certification for the DYNAMIKA™ platform. This is the infrastructure of a company that has been through FDA inspections and regulatory audits and has emerged with its data quality intact.

Clinical trial imaging outsourcing decisions made on the basis of price alone — choosing the cheapest imaging vendor without interrogating their compliance infrastructure – have a well-documented tendency to create expensive problems downstream. The cost of a failed trial or a regulatory complete response letter vastly exceeds any savings from underpriced imaging CRO services. Clinical trial imaging outsourcing is a risk management decision as much as a resource decision: the right imaging CRO selection protects the integrity of the primary endpoint, and the wrong one can compromise it irretrievably.

The imaging CRO RFP (request for proposal) process has become more sophisticated as sponsors have accumulated experience with what can go wrong in imaging trials. But many sponsors still issue imaging CRO RFPs that focus primarily on price and turnaround time, rather than the capabilities that determine whether the imaging data will hold up under regulatory scrutiny. Imaging CRO selection is consistently underweighted as a strategic decision in early trial planning and consistently overweighted in importance when a trial is struggling mid-study and sponsors wish they had chosen differently.

What should a rigorous imaging CRO RFP actually ask?

Regulatory compliance and audit history. The imaging vendor should be able to demonstrate compliance with 21 CFR Part 11 for all electronic records and electronic signatures. Ask for documentation of SOC 2 Type II audit status, ISO 13485 certification, and experience with FDA and EMA inspections. An imaging core lab that has never been through a regulatory inspection is a different risk profile than one that has been inspected repeatedly and consistently passed.

Imaging charter development expertise. The imaging CRO RFP should ask for examples of imaging charters developed for comparable indications. Imaging charter development is a specialized skill – a charter written by someone unfamiliar with your therapeutic area will likely create ambiguities that generate protocol deviations. Ask specifically whether the imaging CRO has written charters applying RECIST 1.1, iRECIST, volumetric tumor assessment, RANO criteria, or whichever criteria your trial requires.

Centralized image management infrastructure. How does the imaging core lab receive, quality-check, and store images? What happens when a site sends images that don’t meet the protocol’s quality requirements? What is the process for managing queries, resolving issues, and ensuring that the image dataset is complete at database lock? Centralized image management is the operational foundation that everything else rests on – it deserves detailed scrutiny in the imaging CRO RFP.

Blinded independent central review workflow. For any trial where BICR is expected by regulators, the imaging CRO’s process for maintaining reader blinding, managing conflicts between dual readers, and conducting adjudication must be fully documented. Ask for the standard operating procedures. Ask how reader training is conducted and how reader qualifications are documented. BICR is only as good as the process behind it.

Site qualification imaging capabilities. Does the imaging CRO have a formal site qualification imaging process that tests scanner performance against the imaging charter’s requirements before patient enrollment begins? Site qualification imaging is consistently underinvested in by sponsors who are eager to open sites quickly – and consistently regretted when unqualified sites generate unusable data.

AI and quantitative imaging capabilities. If AI imaging endpoints or radiomics are part of your trial’s scientific strategy, ask explicitly how the imaging CRO has validated its AI tools, what regulatory submissions have included AI-derived imaging data, and how the AI outputs integrate with the blinded reading workflow.

Imaging trial rescue experience. No sponsor plans to need imaging trial rescue services. But imaging trials go wrong – protocol deviations accumulate, imaging charter ambiguities create inconsistencies, site compliance degrades mid-study. The imaging CRO that has rescued underperforming imaging trials has operational knowledge that cannot be acquired from textbooks. Ask for it. An experienced imaging CRO can perform imaging trial rescue by auditing the existing imaging dataset, identifying which timepoints and sites are evaluable, retroactively harmonizing acquisition parameters where possible, and building a defensible analysis plan for the data that does exist. Imaging trial rescue is not a last resort – it is a core imaging core lab service that separates full-service imaging CRO providers from those with limited operational depth. Sponsors who ask about imaging trial rescue capability in the imaging CRO RFP – before they need it, will be better positioned to respond if problems emerge.

Therapeutic area depth. An imaging CRO with deep oncology experience and limited musculoskeletal expertise is not interchangeable with one that has built both capabilities over decades. Oncology imaging clinical trial requirements differ substantially from rare disease imaging requirements, which differ from neuro imaging requirements. Match the imaging vendor to the therapeutic context. For rare disease imaging specifically where imaging CRO selection determines whether a novel MRI endpoint will survive regulatory scrutiny, the sponsor should ask to review publications, regulatory submission experience, and specific rare disease imaging trial case studies before making a decision.

Global site network and support infrastructure. How many countries can the imaging vendor qualify sites in? What languages does their site support team operate in? What is the escalation path when a site issue cannot be resolved at the first level of support? For global Phase III trials, these operational details determine whether timeline commitments are realistic.

Kubassova has said publicly that the imaging CRO selection decision is underappreciated as a strategic choice. Companies spend months selecting their primary CRO and minutes selecting their imaging partner and then discover, mid-trial, that the imaging partner’s quality system cannot support a regulatory submission. The selection criteria outlined here represent the accumulated lesson of two decades of watching that pattern repeat.

A well-structured imaging CRO RFP will also probe the imaging vendor’s approach to iRECIST and mRECIST alongside standard RECIST 1.1. Immunotherapy and hepatocellular carcinoma programs are now commonplace in development pipelines, and a clinical trial imaging vendor that handles iRECIST confirmation scans or mRECIST viable tumor measurement as an afterthought. rather than a designed workflow will create reader training problems and adjudication inconsistencies. The imaging CRO RFP should request reader training materials and concordance data for iRECIST and mRECIST specifically, not just RECIST 1.1.

The trajectory of clinical trial imaging over the next decade is visible in the trends that IAG and its peers are already navigating.

Quantitative Imaging clinical trials will become the norm rather than the exception in inflammatory and metabolic disease. As quantitative MRI biomarkers – T1 rho, T2 mapping, MR elastography, DCE-MRI permeability gain regulatory acceptance as surrogate endpoints, trials that previously required tissue biopsy for efficacy assessment will be able to use non-invasive quantitative MRI endpoints instead. The same applies for trials that use X-ray and CT; quantitative measurements rather than assessments will be prefeed as they allow for shorter more informative trials, with  smaller patient cohorts.

This is better for patients, faster for trial execution, and creates enormous demand for imaging core lab services with expertise in quantitative analysis. The imaging CRO services required to support quantitative MRI, CT, X-ray, DXA, PET and Ultrsound clinical trials are substantially more complex than those required for standard reading requiring scanner harmonization, sequence standardization, phantom calibration, and analysis pipelines that are validated to the regulatory standard required for primary imaging endpoints.

Multiparametric MRI – the simultaneous acquisition of multiple MRI contrasts to characterize tissue in multiple dimensions is already standard in prostate cancer staging (PI-RADS) and is moving into trial use across multiple oncology and non-oncology indications. Multiparametric MRI creates richer datasets and more sensitive imaging endpoints, but also more complex analysis requirements that require specialized imaging CRO capabilities. As multiparametric MRI becomes a routine tool in oncology imaging clinical trials for prostate, liver, and brain tumors, the imaging core lab’s ability to manage multiparametric MRI data across diverse scanner platforms and reading environments becomes a competitive differentiator.

Whole-body MRI is emerging as a staging and response assessment tool in multiple myeloma, metastatic prostate cancer, and other disseminated malignancies replacing or complementing PET imaging clinical trial protocols in some settings. The IMWG myeloma criteria are evolving to incorporate whole-body MRI findings. PET imaging will not disappear, its role in measuring metabolic activity and therapy response in hematologic malignancies, rare disease imaging studies, and neurodegenerative programs remains essential but the imaging CRO must increasingly manage hybrid workflows that combine PET imaging data with anatomical MRI in the same analysis.

AI imaging will continue its advance from research to regulated clinical endpoint. The path from a promising AI biomarker to an FDA-accepted imaging endpoint is long and requires substantial analytical validation, reader study data, and regulatory engagement but it is a path, and companies are walking it. IAG’s partnerships with Takeda, Ferring, Picture Health, and Better Medicine are, in part, investments in being positioned when AI imaging endpoints become standard rather than experimental.

The competitive landscape for imaging CRO services will consolidate. Smaller imaging core lab operations that lack the compliance infrastructure, therapeutic breadth, and technology investment to compete at the Phase III global trial level will struggle to maintain sponsor relationships as trial complexity grows. The premium will go to imaging CRO services providers with documented regulatory performance, deep therapeutic expertise, and genuine AI capabilities not to those with the lowest line-item cost.

Tumor response assessment is evolving in parallel. The combination of volumetric tumor assessment, radiomics-derived biomarkers, and AI-powered tumor response assessment tools is beginning to provide a richer picture of treatment effect than RECIST 1.1 alone and regulatory agencies are increasingly open to these approaches when the validation data supports them. An imaging CRO that has invested in tumor response assessment methodology beyond conventional linear measurement will be better positioned to support sponsors seeking sensitive, early-response imaging endpoints. IAG’s work on volumetric tumor assessment and AI-powered tumor response assessment across hundreds of oncology trials represents exactly this kind of methodological investment. Imaging CRO selection for next-generation oncology programs will increasingly favor partners with this depth of tumor response assessment capability.

Dr. Olga Kubassova has, in the nineteen years since founding IAG, co-authored more than 120 publications, won four national and international Entrepreneur of the Year awards, been become Ernest Young Entrepreneur of the Year finalist, was named to the LDC / Times Top 50 Entrepreneurs in the UK in 2024, completed a Goldman Sachs mini-MBA with Oxford Said Business School, participated in the NY Venture Fellows program, been elected a Scholar of both the Royal Society of Medicine and the Royal Academy of Engineering, and served as a scientific adviser to the UK government and EU funding bodies.

These biographical details are not incidental. They trace the arc of someone who built a company on a scientific insight and then built the business, institutional, and scientific infrastructure around it to make the company permanent.

What Kubassova saw in that Leeds Computer Science lab – that the pharmaceutical industry needed a genuinely specialized imaging CRO, not a general CRO with imaging capabilities bolted on has been vindicated by nearly two decades of trial outcomes.

The companies that invested in rigorous central imaging review, proper imaging charter development, validated imaging endpoints, and compliant blinded independent central review produced imaging datasets that supported regulatory approvals. The ones that treated the imaging core lab as a commodity service and chose on price alone learned the lesson more expensively.

The breadth of IAG’s imaging CRO services today reflects how far that founding insight has been developed. Central imaging review is just the beginning. Site qualification imaging, imaging charter development, AI imaging endpoints, radiomics integration, imaging trial rescue, multiparametric MRI analysis, quantitative CT biomarker development, rare disease imaging endpoint design, PET imaging protocol management – each of these capabilities represents years of investment in expertise, validated workflows, and regulatory track record. No sponsor needs all of these capabilities on every trial. But knowing that the imaging CRO they have selected can provide them, at the quality required for regulatory submission, is what makes imaging CRO selection a strategic decision rather than a purchasing decision.

The field of clinical trial imaging has matured enormously since 2007. Regulatory expectations have risen. Response criteria have multiplied and refined. AI imaging capabilities that did not exist when DYNAMIKA™ V.1 was built are now production tools in clinical trials. The clinical trial imaging outsourcing market has grown from a niche specialty into a multi-billion-dollar segment of the global CRO industry.

Through all of it, the core insight remains the same as the one that earned a mathematician from Kazakhstan a £5,000 prize, government grant and a company to build: if you can measure it precisely and reproduce it reliably, you can know whether the drug is working. Everything IAG has built – the platform DYNAMIKA™, the processes, the partnerships, the global site network, the reader network, the regulatory infrastructure is in service of that proposition.

In the next decade, as quantitative imaging endpoints move into mainstream use, as AI imaging becomes a regulated tool rather than an experimental one, as rare disease, auto-immune, oncology and metabolic, obesity programs demand novel imaging biomarkers that have never been validated before, and as the global clinical trial system becomes still more complex, the case for specialized imaging CRO services will only grow stronger.

The imaging core lab is not a back-office function. It is the measurement system at the heart of how medicine proves it works. Choosing the right imaging CRO – one with the regulatory depth, scientific expertise, technology infrastructure, and institutional experience to manage that measurement system is one of the most consequential decisions a drug developer makes.

Image Analysis Group has spent nineteen years earning the right to be that choice.