Periodic Reporting for period 1 - iPROLEPSIS (iPROLEPSIS: PSORIATIC ARTHRITIS INFLAMMATION EXPLAINED THROUGH MULTI-SOURCE DATA ANALYSIS GUIDING A NOVEL PERSONALISED DIGITAL CARE ECOSYSTEM)
Okres sprawozdawczy: 2023-01-01 do 2024-06-30
Psoriatic Arthritis (PsA) is a chronic, progressive, inflammatory disease affecting 1-2% of the general population, while manifesting in up to 30% of people with psoriasis (PsO). The transition from health to PsA is currently untraceable; diagnosis of early PsA is challenging even in PsO patients. Untimely diagnosis is common and contributes to early deterioration of quality of life, also increasing the burden of the multiple comorbidities associated with PsA.
In this vein, iPROLEPSIS aspires to shed light upon the health-to-PsA transition with a comprehensive multiscale/multifactorial PsA model employing novel trustworthy AI-based analysis of multisource and heterogenous (i.a. in-depth health, environmental, genetic, behavioural) data, digital phenotyping of inflammatory symptoms with emphasis on tracking of motor manifestations using smart devices and wearables, novel optoacoustic imaging-based markers of PsA in the skin and joints, and investigation of the role of mast cells in the PsA transition, to identify key drivers of the disease and support personalized models for PsA risk/progression prediction and monitoring as well as associated inflammation detection and severity assessment.
To ultimately advance PsA diagnosis and care, the models will be translated into a digital health ecosystem comprising dependable tools for supporting healthcare professionals in disease screening, monitoring and treatment via quantitative, explainable evidence, and empowering people with/at risk of PsA with tailored insights and preventive interventions based on actionable factors for educated health management. The project will steer its research and development efforts following a trustworthy framework for ethical, lawful, and robust AI, and a user centered co-creation approach based on constant involvement of key stakeholders during the design, development, and testing of the digital health ecosystem, securing successful integration of the latter in the continuum of care.
In this vein, iPROLEPSIS aspires to shed light upon the health-to-PsA transition with a comprehensive multiscale/multifactorial PsA model employing novel trustworthy AI-based analysis of multisource and heterogenous (i.a. in-depth health, environmental, genetic, behavioural) data, digital phenotyping of inflammatory symptoms with emphasis on tracking of motor manifestations using smart devices and wearables, novel optoacoustic imaging-based markers of PsA in the skin and joints, and investigation of the role of mast cells in the PsA transition, to identify key drivers of the disease and support personalized models for PsA risk/progression prediction and monitoring as well as associated inflammation detection and severity assessment.
To ultimately advance PsA diagnosis and care, the models will be translated into a digital health ecosystem comprising dependable tools for supporting healthcare professionals in disease screening, monitoring and treatment via quantitative, explainable evidence, and empowering people with/at risk of PsA with tailored insights and preventive interventions based on actionable factors for educated health management. The project will steer its research and development efforts following a trustworthy framework for ethical, lawful, and robust AI, and a user centered co-creation approach based on constant involvement of key stakeholders during the design, development, and testing of the digital health ecosystem, securing successful integration of the latter in the continuum of care.
Key achievements during the 1st reporting period include the creation of a foundational body of knowledge on target diseases and retrieval of relevant datasets; identification of user needs and engagement with key stakeholders to steer research and co-create digital care tools; establishment of the trustworthy AI framework; development of early versions of digital indicators for detecting inflammatory symptoms; development of the first version of the digital health ecosystem for personalised preventive care; and progress in clinical studies, including preparing protocols and obtaining ethical approvals.
Knowledge Mining, Foundation, and Participatory Design: A comprehensive literature review explored methodologies, research prospects and challenges in detecting and monitoring PsA and PsO and their triggers. Large-scale databases were identified and assessed and a common data model (OMOP) was defined to harmonise retrospective and prospective data. An agile co-creation approach was established, engaging key stakeholders to map user needs and identify core user stories for designing the iPROLEPSIS digital health ecosystem through focus groups, online surveys, and workshops. The iPROLEPSIS trustworthy AI framework was established to govern the project's lifecycle, incorporating European requirements, ethical considerations and responsible AI standards.
Research on PsA Inflammation Drivers and Monitoring Research: Data analytics tools were designed to mine 3 datasets (medical history, lab tests, patient-reported outcomes, and environmental stressors) from existing cohorts to identify triggers of PsA transition and inflammation exacerbation. Early versions of digital biomarkers were implemented to assess inflammatory symptoms, focusing on pain, fatigue, and morning stiffness, using touchscreen typing data, smartphone and wearable sensor data, skeleton tracking from videos, nail images, and lifestyle monitoring. An early fusion approach was designed to integrate dBMs with electronic health records to create dynamic, multiscale, xAI-driven models for predicting inflammation exacerbation.
The iPROLEPSIS Digital Health Ecosystem: The high-level system architecture and technical specifications for the iPROLEPSIS ecosystem were formulated. The miPROLEPSIS study patient app was developed to facilitate research and data collection. AI models for personalised nutrition and physical activity recommendations were developed, along with the biAURA intervention targeting sleep using binaural beats. Early versions of the personalised serious games suite (PGS) for individuals with PsA were also designed and developed. A cloud-based data management infrastructure was established for secure data storage and processing. System monitoring tools were evaluated, and a continuous integration structure was designed.
Clinical Studies: The PsA digital phenotyping and inflammation (PDPID) study protocol was prepared and registered on clinicaltrials.gov study tools and procedures were arranged, and ethical approval was obtained in three countries. A protocol for the mast cells and optoacoustics-enabled microvascular imaging (MOMJI) study was also prepared. Activities were initiated to integrate the inflammation digital biomarkers validation (IDBV) study with the HIPPOCRATES IMI project.
Knowledge Mining, Foundation, and Participatory Design: A comprehensive literature review explored methodologies, research prospects and challenges in detecting and monitoring PsA and PsO and their triggers. Large-scale databases were identified and assessed and a common data model (OMOP) was defined to harmonise retrospective and prospective data. An agile co-creation approach was established, engaging key stakeholders to map user needs and identify core user stories for designing the iPROLEPSIS digital health ecosystem through focus groups, online surveys, and workshops. The iPROLEPSIS trustworthy AI framework was established to govern the project's lifecycle, incorporating European requirements, ethical considerations and responsible AI standards.
Research on PsA Inflammation Drivers and Monitoring Research: Data analytics tools were designed to mine 3 datasets (medical history, lab tests, patient-reported outcomes, and environmental stressors) from existing cohorts to identify triggers of PsA transition and inflammation exacerbation. Early versions of digital biomarkers were implemented to assess inflammatory symptoms, focusing on pain, fatigue, and morning stiffness, using touchscreen typing data, smartphone and wearable sensor data, skeleton tracking from videos, nail images, and lifestyle monitoring. An early fusion approach was designed to integrate dBMs with electronic health records to create dynamic, multiscale, xAI-driven models for predicting inflammation exacerbation.
The iPROLEPSIS Digital Health Ecosystem: The high-level system architecture and technical specifications for the iPROLEPSIS ecosystem were formulated. The miPROLEPSIS study patient app was developed to facilitate research and data collection. AI models for personalised nutrition and physical activity recommendations were developed, along with the biAURA intervention targeting sleep using binaural beats. Early versions of the personalised serious games suite (PGS) for individuals with PsA were also designed and developed. A cloud-based data management infrastructure was established for secure data storage and processing. System monitoring tools were evaluated, and a continuous integration structure was designed.
Clinical Studies: The PsA digital phenotyping and inflammation (PDPID) study protocol was prepared and registered on clinicaltrials.gov study tools and procedures were arranged, and ethical approval was obtained in three countries. A protocol for the mast cells and optoacoustics-enabled microvascular imaging (MOMJI) study was also prepared. Activities were initiated to integrate the inflammation digital biomarkers validation (IDBV) study with the HIPPOCRATES IMI project.
The iPROLEPSIS project aims to achieve results beyond the current state-of-the-art, relying on breakthroughs and innovations from clinical studies. Progress involves reviewing the latest theories, technologies, and best practices for targeted diseases; securing approval for the PsA digital phenotyping and inflammation (PDPID) study; designing and developing the first version of the iPROLEPSIS digital health ecosystem for personalised preventive care; and developing preliminary AI pipelines using existing data. Tangible contributions include ethical approval of a novel clinical protocol in three countries, the design and development of the miPROLEPSIS study patient app, and the publication of five scientific papers on innovative technologies for personalised PsA inflammation prevention. These achievements create a solid foundation for conducting clinical studies and analysing data, promising significant scientific impact.