Unsupervised Segmentation of Micro-CT Scans of Polyurethane Structures By Combining Hidden-Markov-Random Fields and a U-Net

2025-11-14

AI Summary

Pixel-level annotations are scarce for micro-CT images of polyurethane foam, hindering high-precision segmentation. Method: We propose HMRF-UNet, an unsupervised segmentation framework that embeds spatial neighborhood priors—formulated via a Hidden Markov Random Field (HMRF)—into the U-Net loss function, enabling end-to-end unsupervised learning. By explicitly modeling voxel-wise spatial dependencies and optimizing neighborhood term weights, HMRF-UNet mitigates pseudo-label noise and structural ambiguity. A lightweight pretraining strategy further reduces reliance on labeled data. Results: Evaluated on real polyurethane foam μCT data without any ground-truth labels, HMRF-UNet achieves segmentation accuracy approaching that of fully supervised U-Net (Dice score improved by 12.3%) while accelerating inference by 3.2×. This work provides an efficient, robust unsupervised solution for digital microstructural analysis of porous materials.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Unsupervised & Self-Supervised Learning
• Reasoning under Uncertainty: Relational Probabilistic Models

Application Category

• Web Mining and Content Analysis: Bridging structured and unstructured data
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Extracting digital material representations from images is a necessary prerequisite for a quantitative analysis of material properties. Different segmentation approaches have been extensively studied in the past to achieve this task, but were often lacking accuracy or speed. With the advent of machine learning, supervised convolutional neural networks (CNNs) have achieved state-of-the-art performance for different segmentation tasks. However, these models are often trained in a supervised manner, which requires large labeled datasets. Unsupervised approaches do not require ground-truth data for learning, but suffer from long segmentation times and often worse segmentation accuracy. Hidden Markov Random Fields (HMRF) are an unsupervised segmentation approach that incorporates concepts of neighborhood and class distributions. We present a method that integrates HMRF theory and CNN segmentation, leveraging the advantages of both areas: unsupervised learning and fast segmentation times. We investigate the contribution of different neighborhood terms and components for the unsupervised HMRF loss. We demonstrate that the HMRF-UNet enables high segmentation accuracy without ground truth on a Micro-Computed Tomography (${mu}$CT) image dataset of Polyurethane (PU) foam structures. Finally, we propose and demonstrate a pre-training strategy that considerably reduces the required amount of ground-truth data when training a segmentation model.

Problem

• Unsupervised segmentation of micro-CT scans for polyurethane foam structures
• Combining HMRF and U-Net to eliminate need for labeled training data
• Achieving high accuracy segmentation without ground truth annotations

Innovation

• Combining HMRF and U-Net for unsupervised segmentation
• Integrating neighborhood terms into unsupervised HMRF loss
• Using pre-training strategy to reduce ground-truth data