Real tools. Real standards. Real impact. Bigpicture from data to deployment.

“We’ve put something into motion that people have talked about for years, but now it’s real. That’s something to be proud of,” says Jeroen van der Laak, Bigpicture Coordinator and Professor of Computational Pathology at Radboudumc. 

Standardisation as a foundation for everything

One of the big breakthroughs in Bigpicture is the Data Sharing Agreement (DSA), a legal framework signed by 44 partners that makes it possible to share medical data at scale. This agreement enables ethical, GDPR-compliant access to pathology data across countries and institutions. “That alone is quite extraordinary. It means we’ve mapped out and solved a huge number of potential legal and ethical hurdles,” says Jeroen. “Without something like this, we’d still be stuck with small datasets and isolated one-on-one contracts.” 

Building on the DSA, the project is currently developing further agreements for non-beneficiaries, extending the impact of Bigpicture even further into the research ecosystem.

Bigpicture has also helped shape practical standards for digital pathology, such as DICOM compatibility, the Metadata Standard for aligning the annotation of datasets and AI models, and concrete protocols for quality control. Jeroen: “People often talk about standards, but we’re delivering the tools too. We’re not just saying ‘follow this model’, we’re giving people the software to implement it.”

Shaping new policies and education

Bigpicture’s impact reaches far beyond tools and data. The project plays an active role in shaping the future of the field through education and contributing to international standards.

One key contribution is to the European Master in Molecular Pathology program at the University of Nice, where Bigpicture partners introduce students to the fundamentals of AI in pathology. Students from across Europe take part in a six-week internship programme, gaining practical experience at partner institutions and building a European network of future-ready professionals. “So far, we’ve had students from all over Europe. It creates a network of young pathologists who stay connected and stay involved. That’s a really beautiful thing,” says Jeroen.

At the same time, Bigpicture is contributing to the development of a new ISO standard for digital pathology. This formal recognition of the project’s frameworks and workflows strengthens its long-term impact, ensuring that the methods and tools developed within Bigpicture become embedded in broader policy and practice.

A project that evolves with the field

Bigpicture’s infrastructure was never meant to be static. It’s designed to grow and respond to the needs of the field. A clear example is the work we’re doing to develop the first federated Bigpicture node; a major step in decentralised data access. Jeroen: “This is a great innovation. It means that parties that want to share their data without wanting to put it in a central repository can still make it available for research anywhere.” As data sovereignty and privacy requirements shift across Europe, Bigpicture’s architecture is adapting. From centralised hubs to the option of federated access, the project is evolving to meet new realities. Jeroen: “We’re always learning what works best. That’s the strength of this kind of collaboration.”

Changing the way we think

Perhaps the most profound innovation of Bigpicture isn’t a piece of code or a data pipeline, but a shift in culture. The project is helping to redefine what collaboration looks like in medical research. It's changing how institutions think about sharing, standards, and long-term impact. “Can we call it an innovation, preparing the landscape, changing the culture, making people think differently?” Jeroen reflects. “However we want to call our pioneering role in the health industry, it certainly is impact.”

Pioneering change is never smooth. It brings resistance, friction, and moments of uncertainty. Jeroen: “The bumps in the road aren’t signs of failure, they’re what come with being first.” But every agreement signed, every image shared, and every technical tool delivered, we get a step closer to Bigpicture’s vision. And this deeper transformation is already in motion: a new way of working; one that makes room for scale, speed, and openness in AI research.

Preparing the field for impact at scale

Another major step forward is the infrastructure for objective benchmarking and validation. With indirect access and validation tools now in place, the project addresses a growing challenge in biomedical AI. Jeroen: “When a pathologist has three prostate cancer algorithms to choose from, how do they know which one to use? Good benchmarks are critical, and far from obvious.”

And more is coming. As data uploads continue to grow, Bigpicture is getting closer to offering truly large-scale, high-quality datasets for training and validation. The vision ahead is bold, but tangible: a shared infrastructure that empowers researchers, protects data, and accelerates trustworthy innovation. And Bigpicture is delivering the building blocks, step by step.

Innovation in action: these tools make the difference

Bigpicture’s software innovations fall into several key categories, such as quality control, standardisation, and AI tooling. Jeroen: “These are not conceptual ideas or experimental pilots; they are real, concrete tools developed and available for real use. Their purpose is to make data sets easier to compile, curate, upload, to use the data for research and algoritm training, and to make collaboration across institutions technically feasible.”

These are some of the tools developed in Bigpicture and/or adapted to the platforms’ needs:

Quality-focused innovations

Standardisation tools and documents

AI tools