heal, and save lives, without affecting the privacy of people





The European project TARTAGLIA, led by the Spanish multinational GMV, in a public-private consortium of 16 entities, is a new example of technology transfer, in this case in the area of ​​health, to improve treatments (chronic diseases, prostate cancer, Alzheimer’s disease or early detection of diabetic retinopathy), select specific medications, etc.

The ‘background’ in the treatment and analysis of data and in the Artificial intelligence by the Spanish company, as well as its previous experience in systems applied to the sector, endorsed its candidacy to lead the project. It will be financed by the European Union through the Next Generation EU funds and has a budget of more than 7.5 million euros.

“We had very clear objectives -explains Inmaculada Pérez Garro, Director of Digital Health at GMV’s Secure e-Solutions- because we collaborated for more than 20 years in public health units, for example, in the Valencian Community, where infectious diseases were analyzed , working with the big data analytics, of advanced statistics (AI was not applied, in health, to production systems at that time)».

Another previous experience to take into account was that of HARMONY, another emblematic European project in which, in its first phase, “we processed records of people with oncohaematological diseases, to analyze the types of mutations, so that medical specialists could personalize treatments and establish surveillance and precision treatments”. And, more recently, GMV is taking part in the OPTIMA project to improve treatment for lung, breast and prostate cancer with artificial intelligence and Real-World Data.

«In this environment -says Inmaculada Pérez- sharing data was one of the most important problems, since these are collected from different departments of different hospitals, from primary care, from the specialized care environment, etc. to obtain the scientific evidence necessary to develop a personalized, predictive and precise medicine for the benefit of the patient. It is necessary to obtain large datasets to achieve the necessary precision, since the current scarcity of sources produces biases in the algorithms that make it difficult to apply them correctly to different population groups.

The challenge is to develop a methodology for combining and harmonizing that structures, combines and harmonizes the data without affecting the privacy and security of patients, but the benefit can be enormous, since the ‘trainings’ of the Artificial Intelligence models ( IA) by GMV make it possible to ‘share’ data without leaving the ‘security ring’, through a federated network to speed up the application of IA in health care systems, which represents an immense ‘fabric of extremely useful data and information spider’ (in TARTAGLIA «We are going to make our algorithms go to hospitals»).

A ‘federated learning’ knowledge network that, until now, did not exist (it could take up to three months to work with 100 records), on which work will be carried out until June 2024, when the project ends, with the ‘kick off ‘ in the first week of February. In this environment, the public-private consortium has obtained the highest score among the projects submitted to the call for R&D Missions in Artificial Intelligence of the Digital Spain 2025 agenda and the National Artificial Intelligence Strategy. It is made up of, among others, the Vall d’Hebron Hospital, one of the most prestigious in Europe, three regional health systems (La Rioja, Canarias and Galicia), the Barcelona Super Computing Center (one of the four in Europe that has a supercomputer) and centers associated with health centers, as well as SMEs and startups.

Highly applied technology

Among the project’s use cases, GMV highlights the contribution in the use of ultrasound as another of the disruptive contributions of technology to health, to society. It can be applied, for example, by nursing staff or non-specialized health personnel, in ultrasounds, in which an AI algorithm will detect possible incidents.

In future long-distance manned space travel, such as to Mars, it will be necessary to take special care of the health of the crew. Within diagnostic tools, ultrasound is considered the ideal because it is compact and non-ionizing, allowing an immediate image in real time. However, these images are difficult to obtain and require specific training that is difficult for crew members to acquire. Carlos Illana, product manager for GMV’s secure e-solutions, which is leading a project for the European Space Agency to develop technology that facilitates the acquisition of said image, stresses that “the technology under development will make it easier for people with basic training in ultrasound they can obtain ultrasound diagnostic plans so that they can later be evaluated remotely by a specialist. This technology will generate great value also on Earth, as an example of technology transfer developed in the space field.

The federated network platform launched by GMV will, for example, allow hospitals to generate a database of sufficient size to process ultrasound images while maintaining information security, another of the project’s challenges, since, as Illana emphasizes: «Ultrasounds, being operator dependent, are not acquired as regularly as other imaging modalities, such as resonances, for example, which requires long pre-processing and screening phases, which makes it difficult to obtain large databases that generate robust artificial intelligence algorithms of clinical value.

For this reason, applied research such as these carried out in TARTAGLIA favors the democratization of the use of ultrasound, called to be the new stethoscope, and advance diagnoses. For example, a nurse, family doctor or pediatrician can make a ‘screening’ protocol to detect heart disease early. In a recent study with nearly one million patients, it was identified that 38% of heart failures were detected in emergencies and that 46% of patients had already presented symptoms in the previous six months, concluding that an earlier diagnosis could enable more timely, high-value interventions, addressing disparities and slowing HF progression.”

State-of-the-art collaborative environment

As in other areas of the project, GMV will provide coverage to its project partners, not only in the training of models, but also to be able to access the data platforms of institutions that are inaccessible to them individually.

The project is also faithful to the pattern set by the ‘Next Generation’ funds, which suggest taking into account smaller autonomous communities to support them with the allocated resources, and combining their data baggage with that of larger communities. A task of public-private coordination in which GMV contributes its experience in the ICT sector, as a national reference provider of advanced cybersecurity solutions and services in IP networks, mobility applications and ICT applications for Public Administrations and the development of e-Administration.

TARTAGLIA is quite a ‘country project’, as highlighted by the technology company: «We contribute to creating an application and research model that serves the industry and convert Spain into a country that provides innovation, not a consumer of it». And a great financing effort by GMV, which contributes 65% of the budget. To do this, it has also been necessary to overcome legal and administrative requirements (as in the case of the Ethics Committees), very demanding, in a short time, in order to be able to contribute the experience of all those involved.

See them

Leave a Reply

Your email address will not be published. Required fields are marked *