Announcements‎ > ‎

Summary from the CEI activities during CinC

posted Oct 24, 2017, 9:01 AM by Jaume Coll-Font
CinC  went very well from the perspective of ECGI.
There was a sizable attendance of researchers on ECGI and 3 sessions were devoted to the topic.

Around the conferences, CEI organized a series of events starting with a kick off meeting on Sunday morning.
The attendance to that event was about 25 very motivated people.
During that meeting, the organizing committee presented the three hackathon topics that would be covered during the rest of the conference: pre-processing of ECG data for ECGI, estimation of activation and recovery times and ECG pre-processing for AF.
The researchers interested in each topic then gathered in small sub-groups and decided how would they proceed.

Here is a summary of the work done by each workgroup during CinC:

- ECG preprocessing
The objective of this group is to compare the different pre-processing approaches that each researcher uses and determine their impact on inverse solutions.
To that objective participants will process the 4 datasets currently available, 2 donated by Laura Bear --- including a series of paced beats and a sequence of VT that degenerates into VF --- and 2 donated by Jana Svehlikova --- consisting on healthy subject recordings.
Participants in this group will sequentially apply their methods in each stage: filtering, baseline removal, beat segmentation, etc..  The result of this work will be the analysis of the pre-processing under all combinations of approaches in each stage.

During the meeting, this group managed to process the data in the first stage (filtering) by 4 different contributors. These results already show some differences in the solutions obtained when applying MFS to a single beat.

If you want to actively participate in this group, you can join the mailing list at:  ecgi-signal-preprocessing@googlegroups.com or directly visit the google groups website and request joining the mailing list at https://groups.google.com/forum/#!forum/ecgi-signal-preprocessing.

- Processing of atrial signals:

Following last CinC meeting, the CEI working group on atrial arrhythmias has decided to focus on the following actions: (1) promote joint research on pre and post processing of atrial EGMs and ECGs and (2) elaborate a consensus paper on the ECGi in atrial arrhythmias.


In order to follow-up these activities we’ve created a google group so that we can exchange emails. These communications will be posted in the forum in the CEI webpage (https://sites.google.com/site/ecgimaging/workgroups/atrial-arrhythmia). There you can find a first thread with more details about the future actions of the group. 

Please let Maria Guillem <mguisan@itaca.upv.esor Stef Zeemering <s.zeemering@maastrichtuniversity.nl>  know if you’d like to be subscribed to this group. Otherwise you will not receive further email communications regarding this topic but you could follow our activities in the webpage.



- Activation/Recovery times estimation:
The main objective of this group is to accurately determine activation and recovery times from intracardiac electrograms
The progress of the ART (activation and recovery times) group are the following:

Stage 1: Estimation of Activation times in ventricular signals

    • In simulated epicardial electrograms, comparing to exact ground truth  [KIT simulated on EDGAR]
    • In contact-recorded electrograms, comparing to best available ground truth [Bordeaux torso tank with multiple beats]
    • In ECGI-reconstructed electrograms, comparing to best available ground truth [Bordeaux torso tank with multiple beats]

Stage 2 or 3: Recovery times in ventricular signals

Stage 2 or 3: Activation and recovery times in atrial signals

In order to do so, we will  set up a COVALIC "challenge" to be able to post and compare progress online.


If you want to actively participate in this group, please, contact the mailing list at (ecgi-art@googlegroups.com) or directly visit the google groups page at  https://groups.google.com/d/forum/ecgi-art




At the end of CinC, we all gathered in a closing session where we shared the progress of each workgroup. 
After observing the progress in each sub-group, we concluded that pre-processing and activation times should move onto their own CEI workgroups and the activation times processing merge with the current workgroup on atrial signals. We will set-up COVALIC "challences" for each workgroup and continue the progress off-line during the year.

Given the large number of workgroups (5!), we will not pursue new hackathon events in the upcoming CinC meetings. Instead, we will use the Sunday meeting time to decide which aspects of the current workgroups should we work on during the conference. Then we will use the meeting times during the conference to make progress on those aspects,

As a closing note, we reminded that CEI is a collaborative effort and needs the implication of everyone to succeed. To that purpose, we encourage everyone to help organize and propose ideas to the organizing committee. Anyone can reach it at cei-exec@sci.utah.edu
Comments