More information
Description
Our proposed research seeks to develop a predictive medicine application by applying high-performance computing and analytics to clinical, biological, and population data. It addresses not only the technical challenges of personalising cervical cancer screening, but also puts considerable emphasis on the cost-effectiveness and feasibility of population-based implementation of artificial intelligence (AI) in medicine as well as on actual “real-world” validation of AI algorithms. Opportunities for the prevention and treatment of cervical cancer have already made substantial progress. However, there persists an inequality in disease burden and prevention interventions implemented and scaled-up across the European region. Hence, the general objective of our proposal is to use existing registry and health data to (i) personalise cervical cancer screening for the benefit of citizens and society; and (ii) to create algorithms that can deliver personalized cervical cancer screening recommendations. Our project combines multidisciplinary, scientific excellence from medicine [EE, LV, LT, NO], computer science [NO, EE], clinical medicine [EE, LV], innovative public health [LT], and economics [EE, NO] to enable cancer screening to be delivered according to the distinctive, individual risk profile. These complex methodologies will be translated into user-friendly solutions that make the optimal, personalised intervention accessible and understandable to the clinician and to the individual patient. These objectives will be achieved by intensive international collaboration. For the Baltic project team, this opportunity represents a strategic tool for the integration of research within Europe and has been motivated by our desire to achieve greater impact, to advance knowledge and to transfer learning into practical implementation with tangible outputs and outcomes.
Summary of project results
In the prevention of cervical cancer, significant progress has been made, but there are considerable differences in the implementation of cervical cancer-related disease burden and prevention measures among European countries and regions. In this study, the data from national population-based health registries, biobanks, and research studies was integrated to design activities tailored to individual cervical cancer screening needs safely, effectively, and rapidly.
Using registry and health data, individual risk assessments for cervical cancer were provided. The cost-effectiveness and acceptability of the implementation of risk-based screening for both the screening target population and the service providers were analyzed. The integration of all the aforementioned aspects in one research project was possible thanks to the previous and well-planned collaboration among project partners. The project was multidisciplinary, bringing together the necessary fields for innovation: medicine, including clinical medicine and public health, as well as informatics and economics.
In the frame of the project a total of 15 publications have been published/submitted, including 7 joint publications.
The research findings can be applied in routine practice for planning optimal interventions based on individual needs, which are acceptable to both doctors and patients. The contribution of the project is significant in the prevention of cervical cancer and the advancement of precision medicine. Recent achievements, particularly a successful project management, have already begun to increase visibility in the field of cancer screening.
Project-related activities, including publishing, will continue beyond the official project deadline.
Summary of bilateral results
All tasks within the project were meticulously planned and allocated among partners, with each partner contributing their expertise and competences to ensure the successful execution of the project. The Norwegian partners were leading the development and validation risk prediction modelling technique. It was imperative to have a donor project partner in the collaboration, as they brought invaluable competences in artificial intelligence (AI) modelling and prior experience in studying similar data in Norway. Their expertise greatly enriched the project''s capabilities and insights. Estonian partners were responsible for creating and validating data needed for numerical experiments. Cleaned and curated databases were delivered to Norwegian partner. The Estonian partners were responsible for designing data collection studies, overseeing their implementation, and creating analytical datasets from cleaned/cured data. Also, for evaluating the feasibility of predicting cervical cancer and CIN3+ risk using routinely collected nation-wide health data, and cost effectiveness analysis. The work on promoting a personalised, risk-based screening strategy and identify factors that influence implementation of the strategy was led by Latvian partners. Lithuanian partner leaded the work to support knowledge development and transfer for improved integration of Baltic country research and health systems into the international community including donor states and EU countries.The project has received excellent contributions and input from several different expertise areas. In Estonia, the sample of women is based on the Estonian genome donors. In Norway, the Simula research laboratory has contributed to data analysis for Estonian and Latvian data. The project team is committed to pursuing new research grants, both in collaboration with existing partner teams and by forming partnerships with new ones.