mlsanityData to 14 Oct 2026

Deep Structural Causal Shape Models

arXiv 2208.10950 ↗First posted 23 Aug 2022cs.CV cs.LGcited by 11

On the map: LLM reasoning → Clinical and biomedical applications

Abstract (arXiv)

Causal reasoning provides a language to ask important interventional and counterfactual questions beyond purely statistical association. In medical imaging, for example, we may want to study the causal effect of genetic, environmental, or lifestyle factors on the normal and pathological variation of anatomical phenotypes. However, while anatomical shape models of 3D surface meshes, extracted from automated image segmentation, can be reliably constructed, there is a lack of computational tooling to enable causal reasoning about morphological variations. To tackle this problem, we propose deep structural causal shape models (CSMs), which utilise high-quality mesh generation techniques, from geometric deep learning, within the expressive framework of deep structural causal models. CSMs enable subject-specific prognoses through counterfactual mesh generation ("How would this patient's brain structure change if they were ten years older?"), which is in contrast to most current works on purely population-level statistical shape modelling. We demonstrate the capabilities of CSMs at all levels of Pearl's causal hierarchy through a number of qualitative and quantitative experiments leveraging a large dataset of 3D brain structures.

Key facts. Cited by 11 papers in the map (data through 14 Oct 2026). Placed in Clinical and biomedical applications, one of 17 directions in LLM reasoning. Builds on 48 of its 80 references that are in the map. First posted on arXiv 23 Aug 2022.

Builds on

48 of 80 references.

References · 48

  1. 2014Adam: A Method for Stochastic Optimizationcited by 37,240
  2. 2013Auto-Encoding Variational Bayescited by 4,053
  3. 2019A Comprehensive Survey on Graph Neural Networkscited by 2,919
  4. 2015Variational Inference with Normalizing Flowscited by 2,170
  5. 2016Convolutional Neural Networks on Graphs with Fast Localized Spectral Filteringcited by 2,089
  6. 2018Graph Neural Networks: A Review of Methods and Applicationscited by 1,576
  7. 2019Normalizing Flows for Probabilistic Modeling and Inferencecited by 1,323
  8. 2021Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gaugescited by 1,167
  9. 2013Spectral Networks and Locally Connected Networks on Graphscited by 1,102
  10. 2015Fast and Accurate Deep Network Learning by Exponential Linear Units (ELUs)cited by 1,070
  11. 2016Geometric deep learning: going beyond Euclidean datacited by 1,048
  12. 2018Disentangling by Factorisingcited by 616
36 more
  1. 2019Neural Spline Flowscited by 596
  2. 2016Geometric deep learning on graphs and manifolds using mixture model CNNscited by 468
  3. 2015Autoencoding beyond pixels using a learned similarity metriccited by 407
  4. 2018Pyro: Deep Universal Probabilistic Programmingcited by 380
  5. 20193D Morphable Face Models -- Past, Present and Futurecited by 270
  6. 2018Generating 3D faces using Convolutional Mesh Autoencoderscited by 268
  7. 2017AttGAN: Facial Attribute Editing by Only Changing What You Wantcited by 242
  8. 2020Deep Structural Causal Models for Tractable Counterfactual Inferencecited by 211
  9. 2019Causality matters in medical imagingcited by 193
  10. 2019Learning Likelihoods with Conditional Normalizing Flowscited by 172
  11. 2018Isolating Sources of Disentanglement in Variational Autoencoderscited by 167
  12. 2017SplineCNN: Fast Geometric Deep Learning with Continuous B-Spline Kernelscited by 161
  13. 2017CausalGAN: Learning Causal Implicit Generative Models with Adversarial Trainingcited by 152
  14. 2015Gradient Estimation Using Stochastic Computation Graphscited by 121
  15. 2021Counterfactual Generative Networkscited by 111
  16. 2016Learning shape correspondence with anisotropic convolutional neural networkscited by 95
  17. 2020A Systematic Survey on Deep Generative Models for Graph Generationcited by 94
  18. 2019Neural 3D Morphable Models: Spiral Convolutional Networks for 3D Shape Representation Learning and Generationcited by 90
  19. 2019SpiralNet++: A Fast and Highly Efficient Mesh Convolution Operatorcited by 89
  20. 2017Variational Autoencoders for Deforming 3D Mesh Modelscited by 70
  21. 2018Morpho-MNIST: Quantitative Assessment and Diagnostics for Representation Learningcited by 56
  22. 2020Fully Convolutional Mesh Autoencoder using Efficient Spatially Varying Kernelscited by 53
  23. 2017Deformable Shape Completion with Graph Convolutional Autoencoderscited by 50
  24. 2019Learning to synthesise the ageing brain without longitudinal datacited by 42
  25. 2020Invertible Generative Modeling using Linear Rational Splinescited by 40
  26. 2019MeshGAN: Non-linear 3D Morphable Models of Facescited by 39
  27. 2019Single Image 3D Hand Reconstruction with Mesh Convolutionscited by 32
  28. 2019Degenerative Adversarial NeuroImage Nets: Generating Images that Mimic Disease Progressioncited by 21
  29. 2021A Structural Causal Model for MR Images of Multiple Sclerosiscited by 21
  30. 2019DEMEA: Deep Mesh Autoencoders for Non-Rigidly Deforming Objectscited by 20
  31. 2019Explaining Visual Models by Causal Attributioncited by 18
  32. 2019Mesh Variational Autoencoders with Edge Contraction Poolingcited by 14
  33. 2021Learning Feature Aggregation for Deep 3D Morphable Modelscited by 13
  34. 2018Learning distributions of shape trajectories from longitudinal datasets: a hierarchical model on a manifold of diffeomorphismscited by 8
  35. 2021Mesh Convolutional Autoencoder for Semi-Regular Meshes of Different Sizescited by 8
  36. 20213D Human Shape Style Transfercited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2023: 4; 2024: 2; 2025: 4; 2026: 1.

11 papers.

  1. 2023Emerging Synergies in Causality and Deep Generative Models: A Surveycited by 10
  2. 2025Diffusion Counterfactual Generation with Semantic Abductioncited by 10
  3. 2023A Conditional Flow Variational Autoencoder for Controllable Synthesis of Virtual Populations of Anatomycited by 6
  4. 2024Invariant Shape Representation Learning For Image Classificationcited by 5
  5. 2024Applications of interpretable deep learning in neuroimaging: a comprehensive reviewcited by 3
  6. 2025Vector Representations of Vessel Treescited by 3
  7. 2025Probabilistic Modelling is Sufficient for Causal Inferencecited by 2
  8. 2026Generating Chest X-Ray Counterfactuals by Specialising Foundation Image Modelscited by 1
  9. 2023Studying Therapy Effects and Disease Outcomes in Silico using Artificial Counterfactual Tissue Samplescited by 0
  10. 2023MoleCLUEs: Molecular Conformers Maximally In-Distribution for Predictive Modelscited by 0
  11. 2025Shape Modeling of Longitudinal Medical Images: From Diffeomorphic Metric Mapping to Deep Learningcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2023Mesh2SSM: From Surface Meshes to Statistical Shape Models of Anatomy
  2. 2025Integrating Anatomical Priors into a Causal Diffusion Model
  3. 2025Mesh2SSM++: A Probabilistic Framework for Unsupervised Learning of Statistical Shape Model of Anatomies from Surface Meshes
  4. 2023Point2SSM: Learning Morphological Variations of Anatomies from Point Cloud
  5. 2024Weakly Supervised Bayesian Shape Modeling from Unsegmented Medical Images
  6. 2025CAUSAL3D: A Comprehensive Benchmark for Causal Learning from Visual Data
  7. 2026MorphoFlow: Sparse-Supervised Generative Shape Modeling with Adaptive Latent Relevance
  8. 2024Probabilistic 3D Correspondence Prediction from Sparse Unsegmented Images