WebsiteBanner.png

— WHEN —

ESR 5 - Dissemination Activity
May
24
1:00 PM13:00

ESR 5 - Dissemination Activity

On Wednesday, May 24th, Müjde Akdeniz visited Istanbul Bilgi University,where she got the BSc degree from, to give a talk for the students and professors at a seminar series jointly organised by the Departments of Computer Engineering and Electrical& Electronics Engineering.

She introduced the MARCIUS project, disseminated the results from her individual research so far and shared her experiences as a PhD student as part of the Marie Skłodowska-Curie Actions Innovative Training Networks (MSCA ITNs). The seminar ended with a lively Q&A session with great interest from the audience for the MARCIUS project and the EU MSCA ITN programs.

View Event →
ESR4 - Second paper published
Feb
20
8:00 AM08:00

ESR4 - Second paper published

On February 20th 2023, Cristiana Tiago (ESR 4) successfully saw her paper, “A Domain Translation Framework With an Adversarial Denoising Diffusion Model to Generate Synthetic Datasets of Echocardiography Images”, published on IEEE Access Journal. Following her previous work on echocardiography image generation and deep generative models, the published work focuses on developing a pipeline, using Deep Learning methods, to generate 2D echocardiography images and perform domain translation. She used an Adversarial Denoising Diffusion Model to synthesize 2D apical 4CH echocardiography images and perform image domain translation between different scanners and also different vendors.

The paper can be found here for further information.

View Event →
Feb
8
2:00 PM14:00

ESR5 - Paper published

Mujde Akdeniz (ESR 5), together with immense contributions of Claudia Alessandra Manetti (ESR 1) published the paper entitled “Deep Learning for Multi-Level Detection and Localization of Myocardial Scars Based on Regional Strain Validated on Virtual Patients” on IEEE Access Journal on February 8th, 2023. The proposed work serves as proof of concept for detecting myocardial scars on left ventricle (LV) utilizing the LV strain traces. The method is validated on a dataset of virtual patients simulated with the CircAdapt model (based at the Department of Biomedical Engineering, CARIM School for Cardiovascular Diseases, Maastricht University Medical Center). The proposed algorithm can successfully detect the existence of myocardial scars on 305 virtual patients with 98% accuracy. The open access paper can be found at IEEE Explore for the interested reader.

View Event →
MARCIUS at IUS 2022 in Venice
Oct
14
4:30 PM16:30

MARCIUS at IUS 2022 in Venice

During 3 days (October 10th - 13th), Nitin, Paulo, Cristiana and Mujde (ESRs 2, 3, 4 and 5) attended the International Ultrasonics Symposium in Venice, Italy. Both Cristiana and Mujde presented their work during the poster sessions throughout the conference.

Cristiana presented her poster entitled “3D GAN-Generated Echocardiography Images: a 3D Segmentation Application Case”, and Mujde presented the poster of the abstract also submitted to the conference with the title “Automatic Strain-Based Myocardial Scar Detection Using A Convolutional Neural Network Trained With A Virtual Patient Cohort”.

View Event →
ESR4 - Paper published
Sep
19
8:00 AM08:00

ESR4 - Paper published

On September 16th 2022, Cristiana Tiago (ESR 4) successfully saw her paper, “A Data Augmentation Pipeline to Generate Synthetic Labeled Datasets of 3D Echocardiography Images Using a GAN”, published on IEEE Access Journal. The proposed work focuses on developing an automatic pipeline, using Deep Learning methods, to generate labeled datasets of 3D echocardiography images. The data augmentation tool presented on the paper relies on Deep Generative Models to synthesize 3D echocardiography images with corresponding anatomical labels and on using these datasets to train 3D segmentation algorithms.

ESR 6 also contributed to the paper which can be found here for further information.

View Event →
MARCIUS dissemination at MU
Jan
17
5:00 PM17:00

MARCIUS dissemination at MU

Professor Joost Lumens and ESR1 Claudia Alessandra Manetti have been interviewed by the Maastricht University's newspaper 'the Observant Journal'. They had the chance to explain the role of cardiac modelling, in particular the CircAdapt model developed at Maastricht Univerisity, and the application in the MARCIUS project.

Read the full article here (english version) and here (dutch version)

View Event →
ESR 4 - Dissemination Activity
Nov
22
3:00 PM15:00

ESR 4 - Dissemination Activity

On Monday, November 22, Cristiana went to her old university to perform a dissemination activity. At the Faculty of Sciences of the University of Lisbon, Portugal, she gave a presentation about how it is to be involved in a Marie Curie PhD Programme and to show some preliminary results about her research on Deep Learning and Echocardiography.

View Event →
Sep
27
to Sep 30

Summer School 2021

A very special school. The first gathering ever for all teams (ESRs and supervisors) thanks to the ease of COVID restriction. A scientifically rich program with a lot of social activities in the calm and beautiful district of Løten, Norway.

Science:

Artificial intelligence (AI) course by Jurica Šprem

1 - Theory:

  • An overview on the history and beginning of AI.

  • The landmark applications in the field of medical imaging.

  • The challenges and requirements of a successful machine learning algorithm such as domain knowledge, good quality data.

2 - Hands on sessions, Hackathon:

  • ESRs were divided into 3 groups, 2 group in each.

  • The aim of the project was developing an automatic algorithm for the identification of a given echocardiographic view (such as 4CH(4 chamber) , 3CH, 2CH).

  • Over the following couple of days, the groups worked on the given project with the help of Jurica Šprem the results were discussed with Prof. Jan D’hooge with very valuable feed back and action point for further improvements.

  • The ESRs gained a lot of knowledge from this project despite coming from different disciplines. The engineers learned about the different views of echocardiography, how to identify them and what are the cardiac structures in each view, while ESRs from medical and physiological backgrounds gained a good insight in the basics of AI, coding, and commonly used terms in that field.

Poster presentation

  • The ESRs have individually presented their work in the form of a printed poster.

  • The posters had the same layout and were hanged on the on the wall next to each other, which gave the same environment of a professional poster presentation in an international congress.

  • Each ESR presented his/her poster to the rest of the group and got the feedback from the others specially from more experienced attendees like Joost Lumens, Kristen Mc Leod and Jurica Šprem.

Secondment plan

  • Since the COVID restrictions markedly affected travel and mobility, a dedicated session has been held to discuss the secondment plan of each ESR and any needed modification.

Story telling course

  • Tara Philips who is an experienced TEDx speech couch gave a 4-hour course on storytelling, which is a very useful tool for a scientist to present his/her work, to convince attendees with his ideas or to apply for future grants.

    The course had three parts:

  1.   The theory and introduction.

  2.   Physical and mental preparations before giving a speech or a talk.

  3. Individual story exercise: Each ESR has to prepare a 3-minute story using the skills that he/she learned from the course and tell the story to the rest of the team an note the feedback. 

 

Social activities:

  • For a whole week the teams shared the same place, meaning that different types of authentic food have been cooked.

  • Walks in the nature, in Oslo city centre and talks among the team have created a tight social bonding that not only resulted in making long-life friendships but also would have a very strong positive influence on the progress of the work and the results of the whole project.

  • Big Thank you to organizers of the summer school specially for Kristin McLeod.

  • Big Thank you to Christiana Tiago who spent a lot of time preparing quizzes and fun games for the rest of the team, that was very successful.

View Event →
MARCIUS dissemination at UiO
Sep
26
4:30 PM16:30

MARCIUS dissemination at UiO

ESRs 4 and 5, Cristiana Tiago and Mujde Akdeniz , have been interviewed by the University of Oslo (UiO) about their research and about how is it like to be part of a Marie Curie PhD Programme and the final interview piece has been published online at the UiO Department of Informatics (IFI) website.

Read more here.

View Event →
Winter School
Mar
22
to Mar 26

Winter School

The idea with the first winter school was to meet in the mountains and combine some outdoor activities with team building, lectures, and some hands-on projects. Unfortunately, due to the COVID-19 pandemic, this first meeting followed an online format but this didn’t stop the members to travel to the mountains virtually:

marcius.png

There were four modules covered in the Winter School:

  1. Scientific based courses:

    • ‘Computational modeling in cardiology’ Joost Lumens

      Part I: Computation and cardiac modeling

      First, an overview of the main cardiovascular computation models available now in the research was addressed. Second, the focus shifted to the well-known Circadapt model and how it can be used to increase our understanding of cardiac resynchronization therapy (CRT) working mechanism.

      Part II: CircAdapt pratical

      Use of the CircAdapt simulator (http://www.circadapt.org/downloads) to simulate valvular heart disease.

    • ‘Medical grade software design’ Stian Langland
      The lecture was divided into two parts. The first one was dedicated to techniques involved in the design of software such as the SOLID principle and TDD process. The second one focused on SW development with a major focus on the topic of peer review and IEC 62304.

  2. Hands-on project

    The main goal of these projects was to get the six ESRs familiar with a topic outside of their specific research topics but always related to the MARCIUS project in some way.

    The ESRs were divided into pairs and to each team, an individual project was assigned during the first day of the Winter school.

    The projects carried out during the week were the following:

    • Project 1: local morphology – ESR 1 & ESR 2

      Goal: Learn to extract local morphologic descriptors – i.e. curvature – from 2D segmentation results

      Methods: Extraction of 2D images and reference contours from the database CAMUS, use ground truth segmentation to define local curvature, train a regression model to extract curvature from the segmentation mask.

    • Project 2: PW mitral Doppler – ESR 3 & ESR 5
      Goal: Generate synthetic PW mitral Doppler database

      Methods: Use of the CircAdapt model and COLE, generation of moving scatterer cloud in a tube according to CircAdapt motion profile, build virtual Doppler recordings from virtual MV’s

    • Project 3: physiologic ranges – ESR 4 & ESR 6
      Goal: Extraction of a list of new biomarkers from the strain pattern that can be used to predict response to CRT (extraction of strain indices from CRT)

      Methods: Post-processing of the strain data and data and statistical analysis of the biomarkers.

  3. Presentation of the career development plan for each ESR

    Each of the ESR had the chance to present to the supervisors and the other ESRs, his/her own career development plan. The main questions addressed to the candidate were the following: What motivated you to join the MARCIUS program? Where do you see yourself in 5 years? What do you think you need to do to get there (skills, experience needed, etc.)?

    This workshop leads to meaningful interactions between the supervisors and the ESR, especially when comparing the experience in academia to the one in the company.

  4. Online social activity

View Event →
Oct
13
to Oct 14

ESRs Introduction Meeting

Early Stage Researchers Introduction meeting took place on October 14th, 2020.

The event started earlier on October 13th for team members in Oslo with the Team Building day. A whole day spent walking around the lovely city, sightseeing, enjoying the best food!

1.jpg
12.jpg
4.jpg

Mission accomplished: We proved that we know ‘something’ about Artificial Intelligence at the Escape Games Oslo!

3.jpg

After a full day virtual meeting on October 14th, teams at different sites had the chance to introduce their members, where every researcher touched upon their individual backgrounds, interests and how these will contribute to the Marcius project.

View Event →