Meet the partners

What is your role in the AIDAVA team at North Estonia Medical Centre?
I am the project manager and coordinator at the IT department (called product lead in our institution) for AIDAVA and also the contact person for all related questions that may occur during the project. I am also Workpackage 1 lead for the whole project.

What are you currently working on within the scope of AIDAVA and what are your main goals and objectives in the project?
I am currently finishing up the study protocol for the project with our partners that will form the basis for all the future work: including getting ethical approvals, recruiting patients, onboarding data, testing the prototype etc.

Why do you think, the AIDAVA research is important/will make a difference?
When successful, AIDAVA will help to reduce the amount of work institutions need to put into cleaning patients data in order to create good useful data for research and developments. It will also provide the patients the opportunity to be in control of their own data and involve them in the data management process which is still lacking in many EU countries.

What motivates you, personally?
Desire to learn, opportunity to communicate with people from all over Europe, from different countries and backgrounds. In addition to that, the feeling of creating something truly innovative and useful for a very wide audience.

Fun/personal fact

One of my favorite things to do when not doing research is do tricks on trampolines and renovate/flip old houses.

What is your role in the AIDAVA team at INSTITUTION?
I am the Deputy Team Leader for Task 5.1, Novel DL/NLP tools for text-based content.

What are you currently working on within the scope of AIDAVA and what are your main goals and objectives in the project?
We are tasked with creating novel deep learning and machine learning tools to extract relevant clinical information from unstructured text. More specifically, we are currently exploring the efficacy and feasibility of leveraging large language models (LLMs) for some extraction tasks. In the coming weeks, we will begin narrowing our approach to training novel multi-lingual models.

Why do you think the AIDAVA research is important/will make a difference?
The project aims to solve a huge issue: to make high-quality personal health data interoperable, AI-ready and reusable at scale and across institutions and borders, with the end results being better clinical care and support for clinical researchers. If you are familiar with the healthcare field, you can easily recognize this achievement as the great white whale. On a more granular level, we aim to solve a lot of challenging problems in the area of NLP within the project as well.

What motivates you, personally?
AIDAVA will improve the effectiveness of clinical care and support clinical research. At Averbis, our vision is to improve the lives of patients through AI. We couldn’t be more aligned. Knowing that AIDAVA will improve patients’ lives is motivation enough, but I’m also motivated by our project partners, who are industry leaders and consummate professionals. Working with them is a pleasure, and their excellence pushes me to work even harder to achieve our common goal.

Fun/personal fact

Right now, I’m reading about the life and philosophy of Marcus Aurelius.

What is your role in the AIDAVA team at INSTITUTION?
I am the Task Leader for Task 5.1, Novel DL/NLP tools for text-based content.

What are you currently working on within the scope of AIDAVA and what are your main goals and objectives in the project?
As a specialist in clinical language analysis, terminologies and ontologies I am leading Task 5.1 together with Kris Collins, and I am particularly tasked with the inclusion of the Averbis tools into the text annotation AIDAV workflow in terms of pre-annotation.

This requires a tight interaction with task 4.3, led by Medical University of Graz (MUG) where I hold a professorship in medical informatics.

Bridging these two activities, I am the main author of an annotation guideline used for the annotation of clinical narratives. On the part of MUG I am also the curator of the German Interface Terminology for SNOMED CT, which is used for the pre-annotation of clinical narratives in AIDAVA.

Why do you think, the AIDAVA research is important/will make a difference?
AI-readiness and language-independent reusability of clinical data, supported by standards like SNOMED CT and FHIR as the basis of a canonical, knowledge graph based canonic representation of a lifelong personal health record.

What motivates you, personally?
I am personally motivated in demonstrating the usefulness and feasibility of semantic standards, together with new NLP paradigms for interoperable health data, both for primary and secondary health data interoperability use cases.

Fun/personal fact

One of my favorite things to do when not doing research is reading, hiking, cycling, and playing the piano. One thing I cannot live without is reading and keeping myself informed about the world in all its aspects.