Probabilistic Point Cloud Reconstructions for Vertebral Shape Analysis
This addresses vertebral fracture detection in medical imaging, offering an unsupervised approach that avoids intensity-based features, though it appears incremental in applying variational auto-encoder concepts to point clouds.
The paper tackles the problem of detecting vertebral fractures from point clouds without supervision by proposing an auto-encoding network that learns to reconstruct only healthy vertebrae, achieving an area-under-ROC curve of >75% on a dataset of ~1500 vertebrae.
We propose an auto-encoding network architecture for point clouds (PC) capable of extracting shape signatures without supervision. Building on this, we (i) design a loss function capable of modelling data variance on PCs which are unstructured, and (ii) regularise the latent space as in a variational auto-encoder, both of which increase the auto-encoders' descriptive capacity while making them probabilistic. Evaluating the reconstruction quality of our architectures, we employ them for detecting vertebral fractures without any supervision. By learning to efficiently reconstruct only healthy vertebrae, fractures are detected as anomalous reconstructions. Evaluating on a dataset containing $\sim$1500 vertebrae, we achieve area-under-ROC curve of $>$75%, without using intensity-based features.