National Centre for HPC, Big Data and Quantum Computing (ICSC), a project funded by the European Union – NextGenerationEU – and by Italy’s National Recovery and Resilience Plan (PNRR) – Mission 4, Component 2.
Objectives
The social burden of neurodegenerative disorders (NDDs) in the elderly is growing worldwide. Early diagnosis and prevention are needed to contain their progression and the related healthcare costs.
The project aims to develop personalized artificial intelligence solutions to determine the risk of developing NDDs based on data derived from multiscale NDD health datalakes, made up of available clinical databases, available disease models, and previously proposed predictive algorithms.
Initial challenge
The prevalence of Alzheimer’s disease and related dementia disorders (AD-NDD), currently affecting 57.4 million people worldwide, is expected to grow to 152.7 million by 2050. The prevalence of Parkinson’s disease and related movement disorders (PD-NDD) is expected to double by 2040, from the current 7 million to 14 million worldwide. Given the significant impact on quality of life, personal and family suffering, and social costs, early diagnosis and prevention represent the most promising approach, with the aim of containing the progression of these disorders and their associated healthcare costs.
Knowing that there is a reasonable risk of developing an NDD within the next 4-8 years could lead to informed decisions regarding lifestyle changes, engagement with new disease-modifying treatments, and prospective assessment of the impact on personal/family organization over the years. Recent scientific findings have confirmed that lifestyle changes can influence NDD risk, while new disease-modifying pharmacological treatments, effective if dosing begins in the early stages of AD-NDD, will already be available in 2024 (and hopefully for PD-NDD as well within the next 2-3 years).
Solution
The project aims to develop personalized artificial intelligence (AI) solutions to determine the risk of developing NDDs, such as AD and PD, in otherwise healthy adults with early prodromal markers.
The systems are based on convergent data derived from multiscale NDD datalakes, consisting of a broad inclusion of public or private (but available) databases and peer-reviewed, evidence-based datasets, as well as previously proposed disease models and predictive algorithms published in specialized scientific articles.
Benefits
The outcome of this project is expected to provide material for a “proof-of-concept” that could generate intellectual property, contributing to ICSC’s impact in the biomedical area covered by Spoke 8.
It will make a unique contribution to training young AI/Data Science specialists in biomedical applications.
IFAB’s role
IFAB contributed to the development and integration of the data infrastructure and analysis tools needed to enable advanced artificial intelligence approaches applied to the prevention of neurodegenerative disorders. In a field where the quality and interoperability of biomedical data are a precondition for any reliable algorithmic development, IFAB’s contribution covered the entire data management cycle, from collection to making data available for predictive models.
The activities were organized around four main areas:
- Data engineering and data integration: collecting, harmonizing, and reconciling heterogeneous information from different clinical and scientific sources, with the goal of building an interoperable information ecosystem ready for analysis.
- Developing systematic procedures for data cleaning, validation, and standardization, essential activities for ensuring the reliability of the AI models trained on them.
- Designing and implementing the datalake architecture needed to host and manage large volumes of multiscale biomedical information, enabling efficient data access and processing by project partners.
- Of particular relevance in the context of biomedical research, generating realistic synthetic datasets to support the development and validation of AI models, helping to mitigate constraints related to privacy and the protection of sensitive data without sacrificing the statistical representativeness needed for training.
Participation in AIND allowed Fondazione IFAB to consolidate advanced expertise at the intersection of data engineering, artificial intelligence, and biomedical research, helping to lay the technological foundations for developing predictive models capable of estimating individual risk of neurodegenerative disorders and supporting personalized prevention strategies.
Partners
Spokes involved
For more information, contact: projects@ifabfoundation.org
