The project proposal must be labelled <Project proposal – Applied ethical AI on Nordic patient records>
|Deadline:||31 January 2023 at 12.00 noon CET|
|Project timeframe:||2,5 years|
In addition to the funding, there will be a co-financing requirement for the winning consortia of at least +50%.
Purpose and context
The purpose of this call for project proposals is to fund the development and demonstration of an ethical AI-based solution, capable of reading patient records across the Nordic countries – across Nordic languages, borders and different medical health record systems. The project shall demonstrate what can be done in the field of ethical AI applied to the responsible use of health data within the current legal framework and changes to the legal framework that occurs during the full project period.
Through this initiative, Nordic Innovation sets out to promote sustainable growth, innovation, entrepreneurship and global competitiveness of Nordic companies.
Large amounts, high quality and diverse data are key to building models that generalize well while at the same time decreasing the risk of bias and can improve clinical efficiency and patient outcomes. While some companies may already have developed and are using AI-solutions on patient records that they have access to, the usefulness of AI and the scope of new innovations increases considerably when it is possible to work across Nordic borders and utilize the full breadth of Nordic health data. This will improve the quality of developed AI-solutions in healthcare and thus unlock new application areas in the Nordic healthcare systems. With the outcome of this call, Nordic Innovation also aims to position Nordic businesses internationally in exporting Nordic ethical and responsible AI solutions applied to health data.
This is one of two calls for proposals launched simultaneously. The other, related call is aimed at demonstrating the innovation potential in the context-rich patient records for innovating new solutions for solving health care challenges.
Together, the two resulting projects aim at propelling the Nordic region as leading in ethical AI and responsible use of health data, by demonstrating the potential for value creation for Nordic businesses and society.
While registry data is often used in research, this is not as common for patient records. Patient records are often only semi-structured and contain written notes. However, in comparison with registry data, patient records potentially give a broader and deeper context than registry data. When mining these large data sets, it would be useful to have an AI solution.
Ethically applied AI/machine learning-methods to patient records have the potential to radically improve healthcare outcomes. Both by improving workflows and by efficiently detecting clinically relevant information hidden in the massive amount of data in patient records, AI holds great potential in improving life expectancy and quality of life for the individual patient. Analyzing this type of medical data at scale could result in novel methods of identifying and treating diseases and illnesses.
There are several barriers (such as technical, ethical and legal) to unlocking the great potential that Nordic health data holds in providing data insights from context-rich patient records. This call is targeted at tackling the technical challenges within the current legal framework and changes to the legal framework that occurs during the full project period.
One of the main technical challenges is the secure access and exchange of data.
For an AI solution applied to patient records to be able to maximize the output and benefit of available data, it needs to be able to function across systems and languages, while adhering to the highest of standards and guidelines concerning ethical and privacy-related aspects.
For the healthcare system, having AI solutions that can interpret and model based on patient records across languages and systems, improves the ability to utilize the information in patient records. The larger pools and higher quality of data strengthen the usability of the developed solutions.
About Nordic Innovation
Nordic Innovation is an organization under the Nordic Council of Ministers working to promote cross-border trade and innovation, located in Oslo. The Nordic Council of Ministers is the official intergovernmental body for cooperation in the Nordic region. With 26 million inhabitants in the Nordic region, the Nordic cooperation is one of the most comprehensive regional partnerships anywhere in the world.
Who is behind the call?
This project is a cross-sectoral collaboration between the two Nordic Innovation programs; Life Science & Health Tech and AI & Data and is supported by the Ministerial Council for Digitalization, which encompasses cooperation between the Nordic countries and Estonia, Latvia and Lithuania.
The Life Science & Health Tech program and the AI & Data program are two of the eight programs at Nordic Innovation launched by the five Nordic Ministers of trade, innovation and industry in 2021. The programs run 2021 – 2024. The programs represent a determined advance towards a stronger and more sustainable Nordic region and support the vision for the Nordic region to be the most sustainable and integrated region in the world by 2030.
The project combines the Nordic Innovation program visions for Life Science & Health Tech and AI & Data; “The Nordics as the most integrated and sustainable health region in the world”, and “The Nordics as the leading region in digitalisation, ethical AI and responsible use of data.” Thus, this project aims to propel the Nordics to become world leading in ethical AI and responsible use of health data.
The aim is to strengthen the Nordic companies' opportunities in the field of life science and health technology to create new products and solutions for the benefit of Nordic citizens.
Who can apply?
A consortium of minimum three project partners from three different Nordic countries. It will be viewed positively with additional project partners from Baltic countries in the evaluation of the proposal. Startups, cluster organizations/eco-systems, SMEs, larger corporations, R&D institutions, public sector actors, NGOs and civil society organizations are all eligible project partners.
The project will:
- Develop and demonstrate a solution consisting of ethical AI-algorithms to read text and images on digital and if possible, analog, patient records across different patient record systems (interoperable), across Nordic borders and Nordic languages.
- Adhere to the protection of privacy of individuals and follow FAIR principles and existing AI guidelines, standards as well as forthcoming EU regulations in the area of AI and healthcare (EU AI Act, EHDS, Data Governance Act etc.).
- The consortium will be required to coordinate a roundtable serving as a formalized stakeholder dialogue twice yearly. This is to ensure relevance and alignment with clinical practice and overall stakeholder interest and developments in the field, as well as knowledge sharing and dissemination of project results. Members of the proposed roundtable will be decided in collaboration with Nordic Innovation. In preparation for the roundtable the consortium must submit a project status report min. 1 week before the scheduled meeting to participants.
- Events, dissemination and communication of project results.
- Barrier assessment which will be made publicly available: The consortium must undertake a barrier assessment covering especially; the technical, societal and environmental barriers to the implementation of the solution as well as an impact assessment of the resulting solution.
- Business case for the developed solution.
Project partners must be eligible for de minimis state aid under Regulation 1407/2013 or 360/2012 (de minimis Regulations) of 18 December 2013.
Funding can only go to Nordic and Baltic partners, with the work being undertaken in Nordic and Baltic countries. Contributions (financial or in-kind) from a non-Nordic or non-Baltic partner cannot count as co-financing towards the co-financing requirements of the project.
Selection of winning consortium and project proposal
Nordic Innovation will select the project proposal and the consortium behind it that Nordic Innovation deems best suited to deliver on the tasks described in this call.
The project proposals that meet the conditions set out in this call will be examined and evaluated by an expert panel. The expert panel consists of a minimum of three external experts and two advisors from Nordic Innovation. Each expert/advisor will individually evaluate each proposal based on the assessment criteria specified in the call text and rank them, with the best proposal ranked highest. All applicants that submit a project proposal will receive a short, written feedback from the expert panel.
In the evaluation, the weight of the three main criteria are as follows: Excellence 40%, Impact 20% and Implementation 40%.
- The overall quality of the proposed project idea and project plan
- Novelty of the resulting AI solution
- How well the proposed project utilizes the context provided in the patient records (relative to using e.g. only patient registry data)
- Quality of the proposed dataset(s): To what extent the proposed datasets exhibit large enough volume and scale to get results that are interesting, clinically valid, and scalable and replicable. How well do the proposed datasets to be used in the project represent the application area and Nordics
- Commercialization: To what extent does the consortium exhibit the experience, capability, capacity and complementary competencies and a plan for commercializing (implementing) and scaling the resulting solution
- Additional Languages: It will be looked upon as favorable if the AI solution is made with more than two Nordic languages
- Additional patient record sets: In the overall evaluation, it will be viewed positively with additional patient data sets, these can be from either other Nordic countries or from a Baltic country
- Additional consortium partners: In addition to Nordic partners, it will be viewed positively with additional partners from Baltic countries
- Closed-Source vs Open-Source: In the assessment of the proposals, it will be viewed positively if the AI solution is developed as an open-source solution
- User benefits and societal impact: The user benefits and the societal positive impact of the AI solution
- Additionality to exciting projects and initiatives: How well the proposed project relates to existing projects, regulations and initiatives on regional, national, Nordic and EU/global level in related areas (e.g. healthcare data, AI)
- The basis of the business case for the developed solution
- Feasibility of access to the dataset(s): Does the consortium have legal access to or is able to credibly assure access and the right to use a sizable number of patient records to be utilized in the project from a minimum of 2 Nordic countries and in a minimum of 2 Nordic languages?
- The consortium composition: How well do the consortium members complement each other and add value through cooperation across the Nordics? Are the consortium members relevant to the project and is it clear what the consortium members add to the consortium?
- Continuous mapping and assessment of barriers: An important deliverable and part of the project is to undertake a mapping and provide an account of the barriers/bottlenecks and learnings under the project. Here we are referring especially to the legal, technical, societal, economic and diversity/equity & inclusivity and to the extent possible, environmental aspects. How will this be ensured?
- Dissemination of knowledge: How the consortium plans to disseminate and present the project’s results
- Overall feasibility and quality of the proposed project plan – involving risk evaluation, mitigation and contingency plan. How well does the proposed project plan match the deliverables?
- Cost structure – Do the contribution to the project and the proposed received funding and input of co-funding look reasonable?
The project proposal shall be in English and no more than 12 pages long including illustrations and pictures. The project proposal shall be submitted in one pdf file. CVs of core consortium members can be submitted as an appendix to the project proposal but must be submitted in the same PDF file as the project proposal. Please number the pages in the PDF. The budget template can be submitted separately.
In order to receive the budget template, please contact: R.malmborg@Nordicinnovation.org
The project proposal must be valid until eight weeks after the application deadline. Signed letters of commitment from all participating partners shall be included.
By the end of the project, a formal report is to be written up and sent to Nordic Innovation, describing both results and learnings.
Application deadline and important dates
Deadline: 31 January 2023 at 12:00 CET.
The project proposal must be submitted electronically to all three mail addresses:
The project proposal must be labelled <Project proposal – Applied ethical AI on Nordic patient records>.
Top project proposals having the highest score on the award criteria may be invited for a digital interview with the evaluation panel and/or to provide answers to potential questions from the evaluation panel during weeks 8-9 in 2023.
Dialog and interaction with Nordic Innovation during the project period will be agreed to in more detail in the beginning of the project period.
|Call is open:||4 November 2022|
|Deadline for asking questions:||2 December 2022|
|Deadline for submission of proposal:||31 January 2023 at 12.00 CET|
|Evaluation process:||31 January - 7 March 2023|
|Potential interviews with top proposals:||Week 8-9 2023|
|Choice of project partners notified:||No later than 7 March 2023|
|Contract award:||No later than 15 March 2023|
Please note that the dates are subject to change.
The submitter of the project proposal is responsible for ensuring that all questions, requirements and clarification points are answered/highlighted and documented in the project proposal. Project proposals and other information received after the application deadline will not be considered. The Applicant carries the risk of errors or delays in the submission/ sending of the Project proposal.
Nordic Innovation may refuse all project proposals if the result of the evaluation process gives a reasonable ground for doing so. If all the project proposals are refused, Nordic Innovation will notify those who have submitted proposals of the refusal.
The decision about to whom the contract is awarded will be published on Nordic Innovations’ website.
Please note that Nordic Innovation’s Standard Terms and Conditions apply to this call.