# Semantic Latent Space Regression of Diffusion Autoencoders for Vertebral Fracture Grading

arXiv: https://arxiv.org/abs/2303.12031 · HTML: https://mlsanity.com/paper/2303.12031/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 22 of its 30 references that are in the map.
- First posted on arXiv 21 Mar 2023.

## Builds on (most cited first, 22 of 22)

- [Denoising Diffusion Implicit Models](https://mlsanity.com/paper/2010.02502/) (2020)
- [Analyzing and Improving the Image Quality of StyleGAN](https://mlsanity.com/paper/1912.04958/) (2019)
- [Designing an Encoder for StyleGAN Image Manipulation](https://mlsanity.com/paper/2102.02766/) (2021)
- [Diffusion Autoencoders: Toward a Meaningful and Decodable Representation](https://mlsanity.com/paper/2111.15640/) (2021)
- [Synthetic Data Augmentation using GAN for Improved Liver Lesion Classification](https://mlsanity.com/paper/1801.02385/) (2018)
- [VerSe: A Vertebrae Labelling and Segmentation Benchmark for Multi-detector CT Images](https://mlsanity.com/paper/2001.09193/) (2020)
- [Medical Image Synthesis with Context-Aware Generative Adversarial Networks](https://mlsanity.com/paper/1612.05362/) (2016)
- [Generative Hierarchical Features from Synthesizing Images](https://mlsanity.com/paper/2007.10379/) (2020)
- [A Computed Tomography Vertebral Segmentation Dataset with Anatomical Variations and Multi-Vendor Scanner Data](https://mlsanity.com/paper/2103.06360/) (2021)
- [CheXplaining in Style: Counterfactual Explanations for Chest X-rays using StyleGAN](https://mlsanity.com/paper/2207.07553/) (2022)
- [LARGE: Latent-Based Regression through GAN Semantics](https://mlsanity.com/paper/2107.11186/) (2021)
- [Keypoints Localization for Joint Vertebra Detection and Fracture Severity Quantification](https://mlsanity.com/paper/2005.11960/) (2020)
- [Grading Loss: A Fracture Grade-based Metric Loss for Vertebral Fracture Detection](https://mlsanity.com/paper/2008.07831/) (2020)
- [Interpretable Vertebral Fracture Quantification via Anchor-Free Landmarks Localization](https://mlsanity.com/paper/2204.06818/) (2022)
- [Context-Aware Transformers For Spinal Cancer Detection and Radiological Grading](https://mlsanity.com/paper/2206.13173/) (2022)
- [Unsupervised and semi-supervised learning with Categorical Generative Adversarial Networks assisted by Wasserstein distance for dermoscopy image Classification](https://mlsanity.com/paper/1804.03700/) (2018)
- [GLOWin: A Flow-based Invertible Generative Framework for Learning Disentangled Feature Representations in Medical Images](https://mlsanity.com/paper/2103.10868/) (2021)
- [Compression Fractures Detection on CT](https://mlsanity.com/paper/1706.01671/) (2017)
- [3D Convolutional Sequence to Sequence Model for Vertebral Compression Fractures Identification in CT](https://mlsanity.com/paper/2010.03739/) (2020)
- [Faint Features Tell: Automatic Vertebrae Fracture Screening Assisted by Contrastive Learning](https://mlsanity.com/paper/2208.10698/) (2022)
- [Deep Generative Adversarial Neural Networks for Realistic Prostate Lesion MRI Synthesis](https://mlsanity.com/paper/1708.00129/) (2017)
- [Interpretable Vertebral Fracture Diagnosis](https://mlsanity.com/paper/2203.16273/) (2022)

## Built on by (most cited first, 2 of 2)

- [Enhancing Interpretability of Vertebrae Fracture Grading using Human-interpretable Prototypes](https://mlsanity.com/paper/2404.02830/) (2024)
- [Revisiting Diffusion Autoencoder Training for Image Reconstruction Quality](https://mlsanity.com/paper/2504.21368/) (2025)

## Abstract (arXiv)

Vertebral fractures are a consequence of osteoporosis, with significant health implications for affected patients. Unfortunately, grading their severity using CT exams is hard and subjective, motivating automated grading methods. However, current approaches are hindered by imbalance and scarcity of data and a lack of interpretability. To address these challenges, this paper proposes a novel approach that leverages unlabelled data to train a generative Diffusion Autoencoder (DAE) model as an unsupervised feature extractor. We model fracture grading as a continuous regression, which is more reflective of the smooth progression of fractures. Specifically, we use a binary, supervised fracture classifier to construct a hyperplane in the DAE's latent space. We then regress the severity of the fracture as a function of the distance to this hyperplane, calibrating the results to the Genant scale. Importantly, the generative nature of our method allows us to visualize different grades of a given vertebra, providing interpretability and insight into the features that contribute to automated grading.