Anatomica: Localized Control over Geometric and Topological Properties for Anatomical Diffusion Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Kadry, Karim, Abdelwahed, Abdallah, Goraya, Shoaib, Manicka, Ajay, Chutisilp, Naravich, Nezami, Farhad, Edelman, Elazer
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911286621634560
author Kadry, Karim
Abdelwahed, Abdallah
Goraya, Shoaib
Manicka, Ajay
Chutisilp, Naravich
Nezami, Farhad
Edelman, Elazer
author_facet Kadry, Karim
Abdelwahed, Abdallah
Goraya, Shoaib
Manicka, Ajay
Chutisilp, Naravich
Nezami, Farhad
Edelman, Elazer
contents We present Anatomica: an inference-time framework for generating multi-class anatomical voxel maps with localized geo-topological control. During generation, we use cuboidal control domains of varying dimensionality, location, and shape to slice out relevant substructures. These local substructures are used to compute differentiable penalty functions that steer the sample towards target constraints. We control geometric features such as size, shape, and position through voxel-wise moments, while topological features such as connected components, loops, and voids are enforced through persistent homology. Lastly, we implement Anatomica for latent diffusion models, where neural field decoders partially extract substructures, enabling the efficient control of anatomical properties. Anatomica applies flexibly across diverse anatomical systems, composing constraints to control complex structures over arbitrary dimensions and coordinate systems, thereby enabling the rational design of synthetic datasets for virtual trials or machine learning workflows.
format Preprint
id arxiv_https___arxiv_org_abs_2511_20587
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Anatomica: Localized Control over Geometric and Topological Properties for Anatomical Diffusion Models
Kadry, Karim
Abdelwahed, Abdallah
Goraya, Shoaib
Manicka, Ajay
Chutisilp, Naravich
Nezami, Farhad
Edelman, Elazer
Machine Learning
We present Anatomica: an inference-time framework for generating multi-class anatomical voxel maps with localized geo-topological control. During generation, we use cuboidal control domains of varying dimensionality, location, and shape to slice out relevant substructures. These local substructures are used to compute differentiable penalty functions that steer the sample towards target constraints. We control geometric features such as size, shape, and position through voxel-wise moments, while topological features such as connected components, loops, and voids are enforced through persistent homology. Lastly, we implement Anatomica for latent diffusion models, where neural field decoders partially extract substructures, enabling the efficient control of anatomical properties. Anatomica applies flexibly across diverse anatomical systems, composing constraints to control complex structures over arbitrary dimensions and coordinate systems, thereby enabling the rational design of synthetic datasets for virtual trials or machine learning workflows.
title Anatomica: Localized Control over Geometric and Topological Properties for Anatomical Diffusion Models
topic Machine Learning
url https://arxiv.org/abs/2511.20587