AI-BRAIN: the project uniting artificial intelligence and biobanks to study gliomas kicks off in Udine

AI-BRAIN: the project uniting artificial intelligence and biobanks to study gliomas kicks off in Udine

On Tuesday 7 July 2026, in Udine, AI-BRAIN officially gets under way — a European cross-border cooperation project between Italy and Austria that brings together artificial intelligence and biobanks to accelerate research into gliomas, the most common primary brain tumours. Glioblastoma.it ODV is one of the project’s partners — a source of great pride for us, as well as the natural continuation of a commitment we have been pursuing since 2019.

AI-BRAIN is a project funded by the Interreg VI-A Italy-Austria 2021-2027 programme (code ITAT-11-045). Its full title already says a great deal: “Implementing the study of brain tumours through artificial-intelligence-guided biobanking.” The project began in January 2026 and will conclude in March 2028.

The underlying idea stems from a concrete problem. Gliomas are rare tumours — around 7 cases per 100,000 people each year — and at the same time highly heterogeneous from a clinical and genetic standpoint. Precisely because they are rare and complex, no single centre, on its own, sees enough cases to draw solid conclusions. To carry out serious research, it is necessary to pool biological samples and clinical data coming from different hospitals and centres, in full compliance with ethical and privacy rules.

This is exactly what biobanks do: facilities that collect, store and distribute high-quality biological samples, together with the data that make them useful. AI-BRAIN aims to build a cross-border federation of biobanks dedicated to gliomas and — and here lies the novelty — to use artificial intelligence to make those samples far more valuable for researchers.

It may seem like a project far removed from a patient’s daily life. In reality it touches on a point we know well: the molecular classification of the tumour. As we explained when discussing the genetic test in glioblastoma, we now know that two tumours that look alike can behave in completely different ways depending on their genetic signature — and that treatment choices depend on that signature.

For research on these tumours to move forward, we need well-characterised samples, linked to reliable clinical information and shareable between centres. Every sample that is well described and made available to the scientific community is one more building block: it helps to find new biomarkers, to understand why some therapies work and others do not, and to build the treatments of tomorrow. AI-BRAIN works precisely on this infrastructure — invisible to the patient, yet fundamental. The three pillars of the AI-BRAIN project are:

1. A network of biobanks that speak the same language. Three biobanks — those of the University Health Authority of Central Friuli in Udine, the Oncology Reference Centre of Aviano and the Brain Tumor Biobank of Innsbruck — will work to harmonise procedures, protocols and quality standards (according to the international ISO norms) and to create a shared catalogue of the available samples. In practice: making sure a researcher can find and use the right materials, wherever they are located, avoiding duplication and waste.

2. Artificial intelligence that “reads” medical reports. A large part of the most important clinical information — histology reports, discharge letters, image descriptions — still exists in the form of free text, often in PDF, hard to search and to reuse. AI-BRAIN will develop systems based on large language models (LLMs) capable of automatically and in a structured way extracting the relevant information from these documents, after anonymisingthem. One aspect that is decisive for us: these tools will run on internal servers, locally, without any patient data leaving the facilities or being sent to external services — in compliance with the GDPR and the European AI Act. It is the theme we already discussed in artificial intelligence and glioblastoma: what really changes for patients, applied here to a very concrete task.

3. Multi-omics analyses. The samples collected and well described will be analysed with advanced technologies — from nanopore sequencing (which reads not only the alterations of the DNA but also its epigenetic modifications, useful for an ultra-rapid molecular diagnosis) to proteomics on blood samples — in order to characterise the tumours in depth and identify new therapeutic targets and disease predictors.

AI-BRAIN brings together five partners across Italy and Austria: the University Health Authority of Central Friuli (ASUFC) in Udine as lead partner, the Oncology Reference Centre (CRO) of AvianoFraunhofer Austria Research, the Medical University of Innsbruck and the Trieste-based PLUS Srl, specialised in data and AI. Alongside the partnership there are two associated partners: the University of Udine and, indeed, Glioblastoma.it ODV. Within the project we will take care of communication and dissemination: bringing the information to the right audience — patients, family members, caregivers — in an accurate and accessible form, and making clear why donating biological samples and data is a precious gesture for research and for personalised medicine. A bridge, in short, between those who do research and those who live with the disease every day. It is the role that comes most naturally to us, and it is the reason this website exists.

The kick-off meeting will be held on Monday 7 July 2026 in Udine, at the lead partner’s premises. The day will open with a press conference presenting the project, followed by an introduction to the objectives and expected results, the presentation of the partners and their respective activities, and working sessions dedicated to the different strands of work — biobanking, ethical and legal aspects, the IT/AI platform and multi-omics analyses. In keeping with the programme’s spirit, the proceedings will take place in Italian and German.

One final note, with the honesty we always try to maintain: AI-BRAIN is research, and research takes its time. A new therapy will not arrive by 2028. But it is precisely this kind of infrastructure — solid biobanks, well-organised data, tools that make them usable — that will make better treatments possible one day. As always, what we learn along the way we will report here.


If you are facing a glioblastoma or a brain tumour — you, a family member, a friend — the Speranza e Coraggio Project is here for that. You can write to us at info@glioblastomamultiforme.it or call us at +39 335 1272678.