15 public-private firms look for early symptoms of Alzheimer's or stroke to age well and live better

The director of the project, coordinated by Capgemini, together with Ángel Lanas, head of the Digestive System service at the 'Lozano Blesa' University Clinical Hospital in Zaragoza and the startup Starlab, tell D+I, from El Español, the details of this research that will last three years.

A few days ago the world population exceeded 8.000 billion people. There are more of us because, among other things, advances in nutrition and health have allowed life expectancy to lengthen. All this despite persistent inequalities, the damage that humanity is causing to the climate and poverty in many places on the planet.

In Spain, this life expectancy in the last two decades has increased, in round figures, from 76 to 80 years in men and from 83 to 85 years in women, according to data published by the National Institute of Statistics (INE). . In 2035, men would reach 83 years of age and women 88.

The big question is: how do we age? Because in addition to living longer, we want to do so in the best possible conditions and, above all, without ailments. Early prognosis of conditions derived from aging is essential, in many cases, to continue enjoying life at advanced ages.

These are neurological, motor and degenerative diseases for which there is no cure and the solution is to minimize the patient's symptoms.

“Identifying which parameters allow us to identify typical diseases such as Parkinson's, senile dementia, muscle loss or stroke would facilitate their early detection”explains Fabiola Bermejo, director of Hybrid Intelligence, the artificial intelligence area of ​​Capgemini Engineering in Spain, in conversation with D+I.

To work on the classification of these indicators, the AI4HealthyAging consortium has been formed. of which Bermejo is director. An “ambitious” project whose objective is to develop an early prognosis solution during the follow-up of patients with heart failure and stroke, thanks to the design and construction of artificial intelligence algorithms.

The intention is that this system serves as support in decision making for elderly patients. “Other diseases such as sarcopenia, Parkinson's or hearing loss will also be analyzed as part of the project,” confirms Bermejo.

Unbiased and protected data
The project is now in its first phase, data collection. A critical job for which, in addition to having sufficient information to carry out subsequent analyses, it is essential to make an appropriate selection to avoid bias and protect the privacy of patients.

To guarantee both, Bermejo explains that two transversal work packages have been designed. In one of them, guidelines have been established on how to apply ethics to artificial intelligence, “always taking into account that the data is for the benefit of society.”

In turn, a platform has been created where the entities that provide the data are told how to treat them from a technical point of view. In this way, they ensure that their subsequent integration is as coherent as possible and, in addition, allows the project to be scaled to other collaborators who join in the future.

“At Capgemini we are in charge of validating that there is no bias and recommending diversity in the presence of patients,” the director points out. “But the most groundbreaking thing about this project is that we have used 'federated learning' to protect that data.”

It is an artificial intelligence technique that prevents the data from ever leaving its place of origin, be it a hospital or a technology center, although the results obtained with them could be consulted. “That is, access to knowledge yes, but never to patient data,” he says.

“When you apply artificial intelligence, you are not interested in knowing the patients, you need the data to draw conclusions with which you then make plans”, it states. And he gives as an example the recommendations that institutions make to lead a healthy life, such as exercising 30 minutes a day or eating X pieces of fruit.

Medical records and sensors
The project is made up of a mixed team, with health and technological profiles. “We can reach conclusions with what the data says, but we have to know how to interpret it,” says Bermejo.

One of the members of the working groups is Ángel Lanas, head of the Digestive System service at the 'Lozano Blesa' University Clinical Hospital in Zaragoza, a hospital that participates on behalf of the Aragón Health Research Institute.

From here they are providing the clinical information available 'anonymised' of all patients who have undergone a colonoscopy in the five years prior to the pandemic and images of the endoscopies of those who have participated in the colon cancer screening program. “In total, we are talking about about 44.000 patients,” Lanas details to this medium.

Amongst the main challenges that they have in this institution for the processing of these data appears evaluating and curating the quality of databases. “To achieve this, and focused on electronic medical records written in natural language and with high heterogeneity, we have the experience of ITAInnova (Technological Institute of Aragon) in the processing and structuring of this type of information”, points out Dr. Lanas.

Having a sufficient number of images, applying algorithms to develop tools with predictive capacity or establishing criteria that allow interpreting, based on clinical indicators, the results obtained through artificial intelligence, are also part of the challenges of this project.

For its part, Starlab, one of the startups that are part of the consortium, collaborates with the CSIC and the Los Madroños Hospital in collecting data in relation to Parkinson's and Alzheimer's with which, in the first case, develop algorithms that indicate the probability of suffering from this disease, and in the second, early detection of the first symptoms of cognitive decline.

“We use neurotechnology in the form of state-of-the-art encephalogram (EEG) acquisition devices", easy to use, that allow electrical activity to be recorded with minimal discomfort for healthy and aging subjects," Aureli Soria-Frisch, Director of Neuroscience at Starlab, describes to D+I.

12,5 million European euros
Bermejo describes this project as a “golden opportunity” for all members of this public-private consortium of which a total of 15 entities are part. Along with Captermini, Starlab and the Aragon Health Research Institute, the Technological Institute of Aragon, the Cajal Institute of the CSIC and the Aragonese Institute of Health Sciences also participate.

Also included are the Universitat Politècnica de Catalunya-BarcelonaTech (UPC), Universitat de València, Barcelona Supercomputing Center, Spanish Red Cross, Andalusian Foundation for Research and Knowledge Management in Otorhinolaryngology (FAIGESCO), and the startups and companies Bitbrain, Podoactiva, Technaid and Integra.

The project, which has a budget of 12,5 million euros financed through the Next Generation EU funds, is included within the 'Artificial Intelligence R&D Missions 2021' program of the Ministry of Economic Affairs and Digital Transformation, in addition to being part of the Digital Spain 2025 agenda and the National Artificial Intelligence Strategy.

“In addition to applying technology and innovation in an environment that is beneficial for health, this project also seeks to strengthen the industrial network, and collaboration between large and small companies and public entities”, says the director of the artificial intelligence area of ​​Capgemini Engineering in Spain.

For this reason, its objectives also include the gradual incorporation of more entities into the consortium that enrich the development of the platform. “Everything we are developing is open, it is a contribution to science,” concludes.

Read the article on the original website, clicking on the button below this text

Source: El Español, Invertia. Authorship: Noelia Hernandez

Featured Image: Source Pexels.com, Photo by cottonbro studio