Project II: The iHELP Project
In general iHELP:
delivers a novel personalised-healthcare framework that enables the collection, integration and management of health-related data from various sources (medical records, lifestyle, behaviours, social media interactions) in a standardised structure called Holistic Health Records. The data in the HHRs is analysed using advance AI techniques to draw adaptive learning models that are used to provide decision support in the form of early risk predictions as well as personalised prevention & intervention measures (alerts, behavioural nudges, consultations medications, therapies, screening etc) that are delivered through usercentric mobile and wearable applications. The standardised integration of data and recalibration of learning models allows the development of frugal AI techniques that provide timely decision support to all stakeholders in the value chain, including: medical experts and also policy makers using relevant interfaces
A more detailed description of iHELP The specific focus of iHELP is on early identification and mitigation of the risks associated with Pancreatic Cancer based on the application of advance AI-based learning and decision support techniques on the historic (primary) data of Cancer patients gathered from established data banks and cohorts. This analysis helps to (i) determine key risks associated with Pancreatic Cancer, (ii) develop predictive models for identified risks, and (iii) develop adaptive models for targeted prevention and intervention measures. Based on the identification of key risks and availability of respective models, the project selects high-risk individuals (from hospital records and other sources) that are invited to take part in the pilot activities or digital trials. The digital trials are carried out through user-centric mobile and wearable applications that apply proven usability principles to offer more engaging experience for health monitoring, risk assessment and personalised decision support. Close collaboration between clinical and AI experts focus on drawing decision support from the prevention and intervention models against identified/predicted risks and providing personalised recommendations (e.g. lifestyle changes, behavioural nudges, screening test etc) to the participants in the digital trials. In addition to providing the personalised monitoring, alerting and decision support mechanisms, the iHELP (mobile and wearable) technology solutions help in validating iHELP solutions and raising health related awareness at individual level. The (secondary) data gathered through the mobile and wearable applications (concerning life style, behavioural, social interactions and response to targeted prevention and intervention measures) is integrated with primary data in the standardised HHR format – within a big data platform. Frugal AI-based learning techniques are developed to provide near real-time risk assessment based on the integration and availability of primary and secondary data in the standardised HHR format. The availability of HHRs provide opportunities to validate iHELP outcomes (e.g. improvements in quality of life, reduced risks etc) through advance analytic functions. iHELP solutions also help in policy making by providing decision support and social analysis on the design of new screening programs and new guidelines for bringing improvements in clinical, lifestyle and behavioural aspects of the fight against Cancer
My role I am the task leader of task 2.1 (e.g. a task which is under work package 2) and is about the requirements and state of the art analysis of the project technologies. The main goal of this task is to collect the user and system requirements with regards to personalised (individual) and communal (policy making level) health management. The outcomes of this task will be a key input for the architecture and the research activities of the project. Furthermore, efforts will be given from my side at the development of the decision support tool which is a very important task of the iHELP project. This task will provide a self-service decision support solution (with query building, analytic and visualisation capabilities), which will allow clinicians/medical-experts and policy makers to analyse data in a holistic fashion i.e. for an individual, groups of individuals, across multiple data stores and locations. An essential element of the DSS is to serve the needs of different types of users e.g. clinicians need decision support to prioritise key risks, develop targeted recommendations and perform analytics on historic data (e.g. co-morbidity and medications), whereas both individuals ad clinicians need real-time decision support based on the processing of streaming data (e.g. biosignals) to engage individuals on lifestyle, behavioural and QoL aspects. The iHELP DSS approach is to apply service-oriented principles to allow the composition of DSS functionalities from multiple building blocks (e.g. analytic algorithms, queries, visualisations etc) while reflecting the needs to support data federation - not in terms of actual federation but in terms of federated queries which can be distributed and executed across multiple data stores (i.e. various types) and locations. An additional feature of the DSS suite will be its ability to support data analytics pipelines / workflows given that isolated analytics deliver limited-value analytical outcomes but their combination will provide additional insights. The DSS suite will apply a set of visualization operators support the creation of a workflow of transformations and visualizations over the results of analytical queries. The visualisation operators will utilize graph models to build the relationships between the care model elements, the actions / interventions and their associated impact / effect.
Explore more in the iHELP’s website