MARCIUS: A quicker way of diagnosing heart attacks

IFI’s Mujde Akdeniz and Cristiana Tiago are two of six Early Stage Researchers (ESR's) on new Marie Curie-funded project, bringing together young researchers from around the world in a project bridging medicine, engineering, data science, and technology.

Bildet kan inneholde: halskjede, panne, nese, smil, hud.

PhD students Mujde Akdeniz and Cristiana Tiago

MARie Curie Intelligent UltraSound (MARCIUS) is a multidisciplinary project that brings together a consortium of industrial and academic partners across Europe. The project has received funding from the European Union's Horizon 2020 research and innovation program under the Marie Skłodowska-Curie grant agreement No 860745.

Two of these young researchers are Mujde Akdeniz and Cristiana Tiago, PhD’s at IFI.

Marcius is part of Marie Skłodowska-Curie Actions (MSCA) Innovative Training Networks programme with one industrial and two academic beneficiaries. GE Vingmed Ultrasound AS (GEVU), located in Oslo Forskningsparken is the industrial beneficiary of the project.

The project aims to improve the interpretation of ultrasound examinations, -namely the echocardiography data- of patients with cardiovascular diseases (CVDs). The idea is to simplify and automate the workflow of clinicians and help with the diagnosis. In order to accomplish this goal, researchers from complementary disciplines of Biomedical Engineering, Physics, Computer Sciences, Physiology and Clinical Cardiology are working together. Below is a visualization of the goals and processes defined in the project.

Bildet kan inneholde: produkt, font, parallell, sirkel, kunst.
The project is divided into 3 conceptual phases.

Learn:
ESR 1: Generation of representative virtual patients
ESR 2: Realistic simulation of image data for virtual patients
ESR 3: Deep learning based tissue characterization
Analyze:
ESR 4: Deep learning based cardiac segmentation and tracking
ESR 5: Clinical decision support tools
Apply:
ESR 6: Clinical knowledge collection and validation

– Exciting project to be a part of

Mujde Akdeniz is an ESR5 and a PhD student at IFI, working at GEVU.                        – My individual research focuses on finding common features in echocardiography exams of heart patients to identify heart diseases. I am using Artificial Intelligence (AI) technologies to develop algorithms that can potentially solve this problem. The diagnosis process of the heart patients is highly dependent on the quality of the medical images of the heart, the expertise of the clinician and other measurements calculated based on the image data.

In MARCIUS, we aim to develop automated systems and integrate them into these processes to support the clinicians for their decision making.

We are combining knowledge and expertise from medicine, computer science and physics to build a new technology in clinical cardiac imaging which will potentially help cardiac patients to have access to improved healthcare. Our research and findings have the great potential of evolving into a product that will be used in daily clinical practice, says Mujde Akdeniz.

Cristiana Tiago, an ESR4 and PhD student at IFI. Her part in this big MARCIUS project is to generate echocardiography images using Generative Neural Networks and create an automatic pipeline to perform, for example, view recognition and structures detection/segmentation, using Deep Learning algorithms. This synthetically generated data can later be used in order to perform clinical analysis.

– For me, the whole project is exciting because it combines both my main interests: Machine Learning and medical imaging. During my academic path I've always been interested in working with cardiac imaging, regardless of the imaging modality, and then the interest for Machine Learning algorithms came along when I started working on the processing of these images.

MARCIUS is just the perfect combination between these 2 domains! What makes it so exciting is the final goal of it: generate new medical data to be used during the research and development of medical solutions, Tiago adds.

MARCIUS and partners

There are three  beneficiaries and multiple partner organizations: GE Vingmed Ultrasound is the industrial beneficiary of the project and is the host for two Early Stage Researchers (ESRs)  based in Oslo, Norway. The two ESRs are registered to the PhD program at the Department of Informatics, University of Oslo.

Katholieke Universiteit Leuven (KUL) hosts two ESRs in Leuven, Belgium, Maastricht University (UM) hosts two ESRs in Maastricht, Netherlands and Oslo University Hospital (OUH), Jessa Hospital (JH) at Belgium, and Medaphor (MDP) at UK are the partner organizations.

 

A full list of team members and responsible people from partner organizations can be found at the projects webpage.

 

 

 

Publisert 9. juli 2021 12:07 - Sist endret 9. juli 2021 12:07