The Statistical Atlases and Computational Modeling of the Heart (STACOM) workshop has been running annually at MICCAI since 2010. The 12th edition of STACOM workshop will be held in conjunction with the MICCAI 2021 in Strasbourg, France. The STACOM workshop is aiming to create a collaborative forum for young/senior researchers (engineers, biophysicists, mathematicians) and clinicians, working on: statistical analysis of cardiac morphology and dynamics, computational modelling of the heart and fluid dynamics, data/models sharing, personalisation of cardiac electro-mechanical models, quantitative image analysis and translational methods into clinical practice.

Final Program is available

See Below


27 September 2021

9:00 UTC Introduction
9:10 UTC From Spins to Pictures to Digital Organs and Back

Keynote speaker: Sebastian Kozerke, Professor of Biomedical Imaging, ETH Zurich

Among the diagnostic imaging modalities, Magnetic Resonance (MR) imaging stands out as it offers a wealth of contrast mechanisms. Using bipolar magnetic field gradients, motion of spins can be encoded. Depending on the scale or range of motion, modulations of image magnitude or phase can be detected. Coherent motion, as for example found with blood flow in larger vessels, is readily quantifiable as a phase change in the images while incoherent or stochastic motion, as with diffusing water in tissue, is detectable as image voxel-based magnitude attenuation. While both types of motion are encoded with essentially the same MR experiment, they offer very distinct insights into in-vivo anatomy and function.

Beyond motion-sensitive imaging, metabolic MR imaging offers yet another angle of capturing the complex interplay of mechanisms in-vivo. While MR has profited from the abundance of water in the human body, probing non-proton substrates has turned out very challenging given the low polarization of non-proton nuclei, the low concentration of non-proton-based substrates such as organic molecules and the limited sensitivity of detecting non-proton compounds. With break-through advances in dissolution Dynamic Nuclear Polarization (DNP) technology a new era has started allowing for real-time MR imaging of key metabolic substrates in the in-vivo heart and other organs.

In this presentation, we will discuss the importance of understanding the interplay of substrate supply, substrate metabolism and function of the heart. It will be demonstrated that changes in metabolic turnover present early hallmarks of deranging function as seen for example in heart failure. In the presentation the basic ideas of encoding flow, diffusion and metabolic information will be conveyed and critically discussed. The importance of discriminating between an analog encoding realm versus digital inference will be stressed and set in contrast to common beliefs in voxel-based information as it is currently witnessed with machine learning based image classifications. We will show the role of MR image-guided digital organs for inferring sub-voxel information from in-vivo measurements and will paint a future scenario of patient-specific in-silico diagnostics and prediction including approaches to optimal experimental design of imaging.

See biography of Sebastian Kozerke, PhD.

10:00 UTC Regular papers (each: 2 min presentation, 8 min Q&A)

  1. A bi-atrial statistical shape model as a basis to classify left atrial enlargement from simulated and clinical 12-lead ECGs
    Claudia Nagel, Matthias Schaufelberger, Olaf Doessel and Axel Loewe
  2. The Impact of Domain Shift on Left and Right Ventricle Segmentation in Short Axis Cardiac MR Images
    Devran Ugurlu, Esther Puyol-Antón, Bram Ruijsink, Alistair Young, Inês Machado, Kerstin Hammernik, Andrew King, Julia Schnabel
  3. Characterizing myocardial ischemia and reperfusion patterns with hierarchical manifold learning
    Benoît Freiche, Patrick Clarysse, Magalie Viallon, Pierre Croiselle, Nicolas Duchateau
  4. Unsupervised Multi-Modality Registration Network based on Spatially Encoded Gradient Information
    Wangbin Ding, Lei Li, Liqin Huang, Xiahai Zhuang
  5. In-silico analysis of device-related thrombosis for different left atrial appendage occluder settings
    Eric Planas, Jordi Mill, Andy Olivares, Xabier Morales, Maria Isabel Pons, Xavier Iriart, Hubert Cochet, Oscar Camara
11:00 UTC Break
11:15 UTC Poster session 1

  1. Multi-atlas segmentation of the aorta from 4D PC-MRI: comparison of several fusion strategies
    Diana Marin, Arnaud Boucher, Siyu Lin, Chloe Bernard, Marie-Catherine Morgant, Alexandre Cochet, Alain Lalande, Olivier Bouchot, Benoit Presles
  2. Quality-aware cine cardiac MRI reconstruction and analysis from undersampled k-space data
    Inês Machado, Esther Puyol-Antón, Kerstin Hammernik, Gastão Cruz, Devran Ugurlu, Bram Ruijsink, Miguel Castelo-Branco, Alistair Young, Claudia Prietro, Julia Schnabel, Andrew King
  3. Coronary artery centerline refinement using GCN trained with synthetic data
    Zhanqiang Guo, Yifan Zhang, Jianjiang Feng, Eddy Yang, Lan Qin, Jie Zhou
  4. Novel imaging biomarkers to evaluate heart dysfunction post-chemotherapy: a preclinical MR feasibility study
    Peter Lin, Terenz Escartin, Matthew Ng, Melissa Larsen, Jen Barry, Idan Roifman, Mihaela Pop
  5. Vessel extraction and analysis of aortic dissection”, Hui Fang, Zhanqiang Guo, Guozhu Shao, Zimeng Tan, Jinyang Yu, Jia Liu, Yukun Cao, Jie Zhou, Heshui Shi, Jianjiang Feng
  6. Generating subpopulation-specific biventricular anatomy models using conditional point cloud variational autoencoders
    Marcel Beetz, Abhirup Banerjee, Vicente Grau
  7. Hierarchical multi-modality prediction model to assess obesity-related remodelling
    Gabriel Bernardino, Patrick Clarysse, Alvaro Sepúlveda-Martínez, Merida Rodríguez-López, Susanna Prat-Gonzalez, Marta Sitges, Eduard Gratacós, Fatima Crispi, Nicolas Duchateau
12:00 UTC Lunch break
13:00 UTC M2Ms-2 (Multi-Disease, Multi-View & Multi-Center Right) Right Ventricular Challenge results
15:00 UTC Break
15:15 UTC Poster session 2

  1. Simultaneous segmentation and motion estimation of left ventricular myocardium in 3D echocardiography using multi-task learning
    Kevin Ta, Shawn Ahn, John Stendahl, Jonathan Langdon, Albert Sinusas, James Duncan
  2. SPHARM-based statistical shape analysis of the tricuspid valve in hypoplastic left heart syndrome
    Jared Vicory, Christian Herz, David Allemang, Hannah Nam, Alana Cianciulli, Chad Vigil, Ye Han, Matthew Jolley, Beatriz Paniagua
  3. Multi-modality cardiac segmentation via mixing domains for unsupervised adaptation
    Fuping Wu, Lei Li, Xiahai Zhuang
  4. Uncertainty-aware training for cardiac resynchronisation therapy response prediction
    Tareen Dawood
  5. Cross-domain artefact correction of cardiac MRI
    Caner Ozer, Ilkay Oksuz
  6. Detection and classification of coronary artery plaques in coronary computed tomography angiography using 3D CNN
    Jun-Ting Chen, Yu-Cheng Huang, Holger Roth, Dong Yang, Chih-Kuo Lee, Wen-Jeng Lee, Tzung-Dau Wang, Cheng-Ying Chou, Weichung Wang
  7. Influence of morphometric and mechanical factors in the configuration of thoracic aorta finite element modeling
    Ruifen Zhang, Monica Sigovan, Patrick Clarysse
16:00 UTC Regular papers (each: 2 min presentation, 8 min Q&A)

  1. Improved AI-based Segmentation of Apical and Basal Slices from Clinical Cine CMR
    Jorge Mariscal Harana, Naomi Kifle, Reza Razavi, Andrew King, Bram Ruijsink, Esther Puyol-Antón
  2. Valve flattening with functional biomarkers for the assessment of mitral valve repair
    Paula Casademunt, Oscar Camara, Bart Bijnens, Hernán G. Morales
  3. Predicting 3D Cardiac Deformations With Point Cloud Autoencoders
    Marcel Beetz, Julius Ossenberg-Engels, Abhirup Banerjee, Vicente Grau
  4. Neural Plaque Angular Characterizer: Automated Quantification of Polar Distribution for Plaque Composition
    Hyungjoo Cho, Hyun-Seok Min, Soo-Jin Kang, Dongmin Choi, Hwiyoung Kim
  5. Mesh Convolutional Neural Networks for Wall Shear Stress Estimation in 3D Artery Models
    Julian Suk, Pim de Haan, Phillip Lippe, Christoph Brune, Jelmer Wolterink
  6. An Unsupervised 3D Recurrent Neural Networkfor Slice Misalignment Correction in CardiacMR Imaging
    Qi Chang, Zhennan Yan, Meng Ye, Kanski Mikael, Subhi Al’Aref, Leon Axel, Dimitris Metaxas
17:12 UTC Prizes and closing


Paper Submission

The STACOM 2021 workshop accepts regular paper submission describing new methods in the following (not limited) topics:

  • Statistical analysis of cardiac morphology and morphodynamics
  • Computational modeling and simulation of the heart and the great vessels
  • Personalisation of cardiac model, electrophysiology and mechanics
  • Quantitative cardiac image analysis
  • Sharing and reusing cardiac model repository
  • Translational studies of cardiac image analysis in clinical practice

The STACOM 2021 workshop will accept 8-page papers (LNCS-Springer format) as regular submissions or challenges. General rules for the submission:

  1. Initial submission must be anonymised and limited to 8 LNCS pages including the References.
  2. Use the LNCS templates/guidelines.
  3. Camera-ready version of accepted papers should include authors’ names & affiliations and can be up to 12 pages

Note that there might be different rules prior paper submission for the challenges. Please refer to specific guidelines from the challenge website. Selected papers will be published in a Lecture Notes in Computer Science proceeding published by Springer (see previous STACOM proceedings).

M&Ms-2: Multi-Disease, Multi-View & Multi-Center Right Ventricular Segmentation in Cardiac MRI Challenge

In this challenge, we invite participants to implement and evaluate advanced approaches based on machine/deep learning for right ventricular segmentation in a multi-disease, multi-view and multi-center setting. A novel aspect of this challenge is the inclusion of long-axis images to help the automatic definition of the basal plane of the RV, which can be confused with the right atrium.