Aging healthily: AI sees age-related diseases coming

IIS Aragón, IACS and ITA collaborate by applying artificial intelligence (AI) to the early diagnosis and prevention of neurodegenerative diseases, sarcopenia and colon and prostate cancer

The passage of time means beginning to accumulate age-related diseases. It is what makes us age. Today, "in a population as aged as ours, we must provide the health system with tools that detect diseases associated with ageing early, since, By addressing them in early stages, which are easier and cheaper to treat, we will provide people with better health, while maintaining an acceptable quality of life., explains Ángel Lanas, scientific director of the Aragon Health Research Institute (IIS) and principal investigator of one of the work packages of the AI4 Healthy Aging project, in which various Aragonese entities and companies participate.

Artificial intelligence (AI) is revealed as a tool that – Lanas is convinced – "is going to radically change the way we work and at a speed greater than we can imagine right now." In collaboration with the Technological Institute of Aragon (ITA) and the Biocomputing Unit of the Aragonese Institute of Health Sciences (IACS), headed by Carlos Tellería, are already obtaining successful results. applying AI to the early diagnosis of two cancers whose frequency increases with age: colon cancer and prostate cancer.

Colon and prostate cancer

The way to anticipate is to get algorithms developed using artificial intelligence to identify which patients are at higher risk "using mainly clinical and analytical data, without the need for invasive tests such as prostate biopsies or colonoscopies," says Lanas. And it works: «The results obtained are being successful», he says. With the project almost finished, "it remains to validate the designed tools and make them available to the health system." For colon and prostate cancer, tools are also being developed to achieve an automatic histological diagnosis and, in the case of the colon, "so that endoscopic images can indicate the histological characteristics of the polyps or lesions found." 

sarcopenia

Many people over 65 years of age see how they lose muscle mass, strength and functionality. What happens to them is called sarcopenia, It leads to a situation of fragility and often goes unnoticed due to the lack of effective screening or early diagnosis tools. Furthermore, the progression of sarcopenia towards frailty creates a vicious circle: the loss of strength and mobility reduces physical activity, which further accelerates the loss of muscle mass. The result: a negative impact on the quality of life of those affected and a greater burden on health systems and the families that care for them.

The way you walk can be very eloquent. By combining AI and different types of sensors and wearable devices, the ITA and the company Podoactiva are developing "a system that, through the analysis of the gait pattern (the way we walk), detects specific characteristics and patterns associated with the presence of sarcopenia," explains Marina Azpíroz, Head of Innovation at Podoactiva.

Older people have participated in the studies; Using advanced sensors, extensive data has been collected on their walking patterns. They have been used to train AI models that analyze the biomechanical characteristics of gait and determine whether or not the person shows signs of sarcopenia. 

As a result of this work, "Models with a reliability of over 85% have been developed"This represents a significant advance in the early detection of this disease," says José Luis Pérez Lasierra, a researcher at the company.

The ITA has been responsible for generating the artificial intelligence models. Rafael del Hoyo, responsible for Big Data and Cognitive Systems, points out that "We have worked hard to understand human movements, through images and motion sensors or accelerometers. AI can "help with the early diagnosis of these diseases, making small signs trigger the alarm and encourage people to visit the doctor," he says. An AI that can hide in a mobile app or in the sensors of a shoe "who are 'listening' at all times."

Another line, with the company Bitbrain, explores the use of AI for early detection and rapid intervention in mild cognitive impairment, dementia and Alzheimer's, based on measures of brain activity during sleep.

An ethical design in the deployment of AI in the healthcare environment

Artificial intelligence (AI) has demonstrated great performance in the task of find complex patterns in large volumes of data. "Since 2014, AI systems have outperformed human experts in multiple areas, thanks to several advances in deep neural networks trained on large data sets using high-performance computing for, for example, drug discovery, prediction of clinical trial outcomes, personalized medicine, analysis of radiological images for early detection of cancer or aneurysms, dermatological testing..." explains Alba de Martino, a researcher at the Aragonese Institute of Health Sciences (IACS).

But, especially when it comes to health, the application of AI must offer all the guarantees. For this reason, the IACS is responsible for ensuring that AI is trustworthy from the design stage in the AI4 Healthy Aging project. This work package, which is entirely cross-cutting and also involves the University of Valencia, the Barcelona Supercomputing Center and the Red Cross, aims to "define a model for regulatory compliance and compliance" a methodology for the ethical evaluation of the technology developed in the project and its application to the different use cases included in the project: heart disease, sarcopenia, sleep, Parkinson's, mental health, cancer and cognitive impairment due to hearing loss."

In the development of AI tools applied to the health field, says the coordinator of this part of the project at the IACS, "it is essential to take into account the potential impact of these practices and techniques on health equity and, in particular, ethical considerations arising from major inequalities in health outcomes due to factors such as socioeconomic status, geopolitical position and ethnic origin." 

The objectives include "ensuring that the developments to be carried out within the project guarantee the fundamental rights of patients and the autonomy and independence of healthcare personnel in the deployment of AI in the health area." The possible ethical and legal conflicts that must be taken into account for the design, development and use of the application are not lost sight of, as well as Privacy and security requirements necessary depending on the organization of the data and the AI ​​procedures used.

Day 10

On December 10th they will be shown at the ITA the latest advances in the use of artificial intelligence to detect and prevent diseases related to aging. The conference entitled 'Artificial intelligence for healthy aging: detection of sarcopenia and mild neurodegenerative diseases through AI', is an event framed within the AI4 Healthy Aging project. The program will especially highlight the role of Aragonese companies such as Podoactiva and Bitbrain in key areas of the project such as: trustworthy AI by design, neurodegenerative diseases and sarcopenia, sleep and prevalent cancers in the elderly. Lines that are part of the different work packages led by Aragonese companies and institutes such as the Technological Institute of Aragon (ITA), the Aragon Health Research Institute (IIS) and the Aragonese Institute of Health Sciences (IACS).  

The project

• Goal Application of distributed artificial intelligence for early diagnosis and treatment of diseases with high prevalence in aging.
• Financing European Union through Next Generation funds. 12,5 million euros within the Artificial Intelligence R&D Missions 2021 programme.
• Execution period Four years, until December 31, 2024.
• Be a partner The consortium, led by Capgemini, is made up of a total of 15 organisations and entities representing the industrial and healthcare sectors, including the Aragon Institute of Technology, the Cajal Institute (CSIC), the Aragon Institute of Health Sciences, the Aragon Health Research Institute, the Polytechnic University of Catalonia-Barcelona Tech, the University of Valencia, the Barcelona Supercomputing Center-National Supercomputing Centre, the Spanish Red Cross, the Andalusian Foundation for Research and Knowledge Management in Otorhinolaryngology, Bitbrain, Podoactiva, Technaid, Inntegra and Starlab.

Source: Heraldo de Aragón