Crowds of people image

Logo

ASSET is an innovation milieu with the aim of developing and evaluating AI algorithms to assess the individual risk of developing autoimmune diseases, such as diabetes.

Goals of the Milieu

ASSET aims to contribute with a personalised prediction and prevention strategy for autoimmune diseases in the society. In particular, we will study how artificial intelligence (AI) techniques can be applied to learn from existing data to identify (I) individuals at risk of developing type 1 diabetes, and (II) individuals with upcoming/newly diagnosed type 1 diabetes that would benefit the most from precision prevention or early intervention with therapeutic approaches.

Icon

The value of prevention

The goal of prevention is to stop disease before it harms the individual.

Icon

Insights through AI

AI allows machines to learn from data to detect patterns and produce insights.

Icon

Understanding complexity

Access to large amounts of data is a prerequisite to derive actionable insights.

Icon

Ensuring the transition

The step from idea to everyday healthcare practice requires systematic changes.

Icon

Finding the right drug

Individual risk should be associated with personalized preventive treatment.

Icon

Closing the loop

Registries provide the ability to systematically collect sufficient data to improve treatment.

Process

The project consists of several iterative steps, including organising the data for models development, which is followed by the development and back-testing of various AI algorithms. Once those steps are completed, opportunities for patient screening will be explored. Here, we will work with principals, regulatory bodies and other relevant actors alongside medical prerequsites.

01

Data Collection

Data collected in the course of the TEDDY study is at the core of model development.

02

Prediction & Development

Based on the available data, models are created and tested for accuracy.

03

Screening

Samples are collected from a chosen population and evaluated using developed models.

04

Putting it to use

Necessary approvals are pursued for use of AI in production of preventative treatments.

Perspectives

Our starting point is a health systems perspective, where we see a gap between the latest technological developments and clinical practice.

Frida Sundberg

MD PhD, Sahlgrenska University Hospital
The value of prevention

At least 45 million persons live with 1 diabetes, a disease that leads to a lifelong dependence on insulin treatment as well as a high risk of serious complications, including increased mortality and raised risks of myocardial infarction, stroke, blindness and kidney failure. The majority of costs related to diabetes is treatment and care of persons with complications.

In ASSET, we evaluate the potential to identify children with an increased risk to develop T1D and match this to treatments that can prevent the progression to overt clinical disease. If we can prevent T1D this would have a significant positive impact for the individual, their families and society as a whole.

Dennis Masich

CEO, MainlyAI
Insights through AI
AI is a set of technologies that allow machines to use existing (big) data sets to derive valuable insights. This can be realized by using so-called machine learning techniques, wherein algorithms are trained on the aforementioned datasets, and are thus able to derive insights from new data not previously encountered.

In the health area, AI techniques have been applied to create new opportunities for personalizing technology-based health interventions. In ASSET, we will build on this existing research to develop AI algorithms for individuals at risk, and individuals that would benefit the most from precision prevention or intervention.

Åke Lernmark

Senior Professor, Lund University
Understanding complexity
Autoimmune type 1 diabetes is a heterogenous disease in which the pancreatic islet beta cells are destroyed by the immune system resulting in insulin deficiency. The disease is triggered several years before clinical diagnosis and is marked by the presence of autoantibodies against beta cell proteins and a genetic component with the strongest known risk attributable to genes that encode the Human Leukocyte Antigens (HLA).

T1D research encompasses several on-going screening cohort studies such as TEDDY, whose data and infrastructure provide invaluable insights into how children at risk are exposed to environmental factors that trigger the disease and how children with islet autoimmunity progress to clinical T1D. The TEDDY and similar cohorts provide data and information to serve as the backbone of AI analysis in the ASSET research environment, facilitating the development of precision prevention strategies.

Hans Winberg

Sectretary General, Leading Health Care Foundation
Ensuring the transition

Experience shows that taking the step from a separate innovation project or trial to implementing an everyday routine in health care is not a trivial task. A main challenge lies in the receptiveness of the healthcare system, i.e. existing organizational structures, regular budgets and rules and regulations in place must allow to endorse new technology and working methods.

In ASSET, we analyze the organizational, economic and legal prerequisites and consequences of applying precision prevention within Type 1 Diabetes in the Swedish health care system. The aim is to proactively address any obstacles that may to hinder or slow down the transition from the project stage to regular health care practice.

Ulf Hannelius

CEO, Diamyd Medical
Finding the right drug

The field of drug development is going beyond the “one size fits all” diagnostic and treatment approach towards a more individualized, precision medicine approach. The driving forces behind this shift are i) the commercial and health economic inefficiencies of the traditional drug development paradigm and ii) the scientific advances that allow us to identify and target new treatments based on inter-individual heterogeneity in pathophysiology of disease.

In ASSET, we will have the unique opportunity to evaluate screening, risk prediction and precision medicine in the setting of type 1 diabetes, setting the stage for a potential national screening program and a new treatment paradigm of autoimmune disease.

Soffia Gudbjörnsdottir

The Swedish National Diabetes Register
Closing the loop

With the advent of novel, preventive and potentially curative treatments, the concept of value-based health, whereby actual health outcomes are measured relative to costs, is more relevant than ever. National health registries such as the Swedish National Diabetes registry, the largest of its kind in the world, provide the ability to systematically and securely collect sufficient data to deliver value-based healthcare.

In ASSET, we have the opportunity to not only understand the individual risk and potential to prevent or delay T1D but also to connect these insights to outcomes in the registries, thereby closing the data loop to facilitate a true ecosystem for precision health.

Asset Partners

ASSET is a collaborative project between the following companies and organizations.

Diamyd Medical

Diamyd Medical develops precision medicine therapies for type 1 diabetes. The diabetes vaccine Diamyd® is an antigen-specific immunotherapy for the preservation of endogenous insulin production. Diamyd Medical also develops the GABA-based investigational drug Remygen® as a therapy for regeneration of endogenous insulin production and to improve hormonal response to hypoglycaemia.

MainlyAI

MainlyAI is a research and technology-based company offering novel AI and ML methods for predictions, knowledge extraction, key performance indicator monitoring and decision support. The company has been successful in applying state of the art methods of privacy-preserving data and knowledge federation in order to speed up and democratize the introduction of AI technologies in industries.

Sahlgrenska University Hospital

Sahlgrenska University Hospital (SUH) is one of Sweden’s largest university hospitals, in close proximity to all areas of education and research. SUH functions as the county hospital for residents of the Gothenburg region as well as being an engine for healthcare and medical development in Region Västra Götaland. The Pediatric Diabetes Clinic at SUH is a EU/SWEET Centre of Reference with a documented high standard of care.

Lund University

Lund University was founded in 1666 and is repeatedly ranked among the world’s top 100 universities. The University has around 44 000 students and more than 8 000 staff base in Lund, Helsingborg and Malmö. The unique disciplinary range encourages boundarycrossing collaborations both within academia and with wider society, creating great conditions for scientific breakthroughs and innovations. The University has a distinct international profile, with partner universities in almost 70 countries.

Leading Health Care Foundation

Leading Health Care Foundation (LHC) is an independent academic think-tank working to promote dialogue about how health and social care can be improved. LHC conducts research and disseminates knowledge that is relevant to the organization, governance, management and transformation of health care systems. LHC have a broad academic network and an active partner network which currently consists of about 30 organizations from different parts of the sector.

The Swedish National Diabetes Register

The Swedish National Diabetes Register (NDR) was established in 1996 with an overall aim to reduce morbidity and mortality, as well as to maximise the cost-effectiveness of diabetes care. It is the largest diabetes register in the world. NDR is an instrument to facilitate such monitoring and to disseminate findings in an accessible, transparent, comparable and timely manner. The register is both a repository of results and an educational tool for improving local quality assessment efforts.

ASSET is an innovation milieu financed by the Swedish Innovation Agency, Vinnova