Activitaty T1.

Development of an imaging-based method for fibrosis localization and quantification in the heart with atrial fibrillation.

Sub-activity T1.1. Development of MRI protocol for image acquisition with the investigation of 10 healthy volunteers and 10 patients with persistent atrial fibrillation.

Sub-activity T1.2. Image processing and method development for fibrosis localization and quantification.

Sub-activity T1.3. Patient data analysis and evaluation of diagnostic protocol.

Results

  • Protocol optimized for regular acquisition of MRI images
  • Functional software prototype for image segmentation
  • Document with description of the imATFIB software architecture

Results dissemination

  • Published papers:
  1. Title: Dynamic Autonomous Image Segmentation based on Grow Cut

Authors: A.I. Marinescu, Z. Balint, L. Diosan, A. Andreica

Conference: European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN 2018), April 25–27, 2018, Brugges, Belgium.

  • Scientific publications in review:
  1. Title: GPU­Accelerated Heart Segmentation

Authors: A. Andreica, Z. Bálint, L. Diosan, A.I. Marinescu

Conference: International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC 2017), September 21–24, 2017, The West University of Timișoara, Timișoara, Romania.

  1. Title: Autonomous 3D segmentation of MRI images based on Grow Cut

Authors: A.I. Marinescu, Z. Bálint, S. Manole, L. Diosan, A. Andreica

Manuscript sent for the Multi-Modality Whole Heart Segmentation Challenge 2017.

  • Conferences
  1. Title: Applications of Cellular Automata to Medical Image Segmentation

Author: A.I. Marinescu

Scientific presentation at the international summer school Doctoral Intensive Summer School on Evolutionary Computing in Optimization and Data Mining (ECODAM 2017); June 19–22, 2017, Iași, Romania.

  1. Titlu: Unsupervised and Fully Autonomous 3D Medical Image Segmentation based on Grow Cut

Authors: Alexandru-Ion Marinescu, Zoltán Bálint, Laura Diosan and Anca Andreica

Conferință: 20th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing – SYNASC 2018, September 20-23, 2018, Timișoara, Romania