Dutch startups applying AI to medical imaging navigate certification and hospital markets

A handful of Dutch companies are tackling the same problem: radiologists and pathologists who assess large volumes of imaging material every day, under time pressure and with the risk of missed findings. From cities such as Rotterdam, Nijmegen and Eindhoven, they are developing AI software that analyses scans for pulmonary nodules, tumours in breast tissue or abnormalities in pathology slides. The technology differs by modality, but the challenges around certification and hospital integration are widely shared.

What sets this cluster apart from generic AI companies is the combination of clinical focus and regulatory complexity. Bringing medical software to the European market requires CE marking as a medical device under the MDR (Medical Device Regulation), a process that takes years and considerable resources. The funding rounds of these companies reflect that: they are large enough to pursue certification, but small enough to feel the pressure of hospital procurement acutely.

What the companies do and which imaging modality they focus on

Aidence (Rotterdam, founded 2015, €13.2M raised) analyses CT scans for pulmonary nodules. Early detection of lung cancer is clinically relevant because five-year survival rates differ markedly between early and late stages. The software targets radiologists running lung screening programmes.

Quantib (Rotterdam, founded 2013, €5M Series A) focuses on MRI and CT analysis, with an emphasis on objectifying radiology findings. Where a radiologist provides a descriptive assessment, Quantib delivers quantitative outcomes that make comparison over time more straightforward.

Screenpoint Medical (Nijmegen, founded 2014, €27M Series B in 2023) develops software for mammography screening. The company raised the largest round in this cluster, which aligns with the scale of population-based breast cancer screening programmes as an addressable market.

Aiosyn (Nijmegen, founded 2021, €4.4M Series A) works on digital pathology, a modality in which tissue sections are scanned and assessed by AI. Pathology is less advanced in digitalisation than radiology, which brings both opportunities and adoption barriers.

CytoSMART Technologies (Eindhoven, founded 2012, €5.2M) sits at the intersection of laboratory equipment and AI: miniature microscopes that enable real-time cell monitoring. The application is closer to laboratory processes than to clinical diagnostics, but the technological foundation overlaps.

Spectro-AI (Enschede, founded 2018, €19.3M Series B) falls somewhat outside this cluster: the company develops autonomous AI for drone inspections in the field and is therefore more of an industrial player than a medical one.

The MDR process as a shared hurdle

Since the European MDR came fully into force in May 2021, stricter requirements apply to AI software classified as a medical device. Software that supports a diagnosis or influences treatment decisions generally falls under Class IIa or higher, requiring a conformity assessment by a notified body.

For startups, this means investing early in clinical validation studies, technical documentation and quality management systems. Aidence and Screenpoint Medical are further along in this process than younger players such as Aiosyn, which was founded in 2021. The funding history suggests that investors are willing to finance these longer lead times, provided the clinical substantiation is convincing.

An additional concern is that MDR notified bodies face a capacity problem: waiting times for assessment have increased considerably in recent years, extending the certification process further. This affects all players in this segment, regardless of how mature their technology is.

Integration into hospital workflows and procurement routes

CE marking provides market access, but no guarantee of adoption. Hospitals in the Netherlands procure medical software through European public tender procedures once contract values exceed threshold amounts. This means startups must compete simultaneously on price, clinical evidence, interoperability and implementation support.

Integration into the radiologist workstation is technically feasible via HL7 FHIR and DICOM standards, but in practice it runs up against the diversity of hospital information systems and the reluctance of IT departments to add new interfaces. Companies such as Quantib and Aidence work with major PACS vendors to offer their software as a plug-in, which lowers the barrier for hospitals.

Screenpoint Medical partly takes a different route: partnering with screening organisations that run population-based programmes. This changes the nature of the procurement process, but the volumes are potentially larger and decision-making is more centralised.

For pathology, where Aiosyn operates, the digitalisation of the department itself is a prerequisite before AI software can be deployed at all. Many Dutch hospitals are in the process of transitioning to digital pathology, but the rollout is slow. This limits Aiosyn's market in the short term, but positions the company early in a structurally growing segment.

Funding and stage: where the companies stand

Funding rounds range from €4.4M (Aiosyn, Series A) to €27M (Screenpoint Medical, Series B). That difference is related to founding year, stage of clinical validation and market size. Screenpoint Medical raised its Series B in 2023, a point at which the broader AI investment market was cooling, making the amount notable.

Quantib has raised a relatively modest €5M for a company founded as far back as 2013. It is possible that the company pursued profitability early or explored alternative growth financing routes. Without further public information, caution is warranted in interpreting this.

CytoSMART's business model falls partly outside the pure software-as-a-medical-device category because it also sells hardware, which makes its capital requirements and growth profile different from pure software companies.

For the broader Dutch and European AI healthcare ecosystem, this cluster illustrates that medical AI is not a fast market. The combination of regulatory requirements, long hospital sales cycles and the necessity of clinical evidence means companies need multiple funding rounds before reaching scale. Investors and policymakers seeking to strengthen this segment face a structural challenge: how to keep promising companies sufficiently well-capitalised to complete the long journey from certification to widespread adoption.