Masked Registration and Autoencoding of CT Images for Predictive Tibia Reconstruction

TL;DR

The paper tackles predicting a patient-specific healthy tibia reconstruction from fractured CT data to aid surgical planning. It introduces a 3D-adapted Spatial Transformer Network to register fractured CTs to a learned tibia prototype and evaluates multiple autoencoder architectures (VAE, VQ-VAE, PCA) to model healthy tibia variation, with a mask-robust extension enabling reconstruction from fracture-masked inputs. Key contributions include the modified STN for global registration, a comparative analysis of bone CT autoencoders, and a data-augmentation strategy to achieve mask-robust predictive decoding. The approach demonstrates high-quality reconstructions even with masked fracture regions, offering potential to improve preoperative planning and bone-restoration outcomes.

Abstract

Surgical planning for complex tibial fractures can be challenging for surgeons, as the 3D structure of the later desirable bone alignment may be difficult to imagine. To assist in such planning, we address the challenge of predicting a patient-specific reconstruction target from a CT of the fractured tibia. Our approach combines neural registration and autoencoder models. Specifically, we first train a modified spatial transformer network (STN) to register a raw CT to a standardized coordinate system of a jointly trained tibia prototype. Subsequently, various autoencoder (AE) architectures are trained to model healthy tibial variations. Both the STN and AE models are further designed to be robust to masked input, allowing us to apply them to fractured CTs and decode to a prediction of the patient-specific healthy bone in standard coordinates. Our contributions include: i) a 3D-adapted STN for global spatial registration, ii) a comparative analysis of AEs for bone CT modeling, and iii) the extension of both to handle masked inputs for predictive generation of healthy bone structures. Project page: https://github.com/HongyouZhou/repair

Figures (2)

• Figure 1: Overview of the proposed methods. (See details in Sec. \ref{['sec:methods']}.) Top left: Training data $D$ of non-aligned CT images. Step 1 trains an STN and prototype jointly for consistent registration. Step 2 trains alternative AEs to provide latent representations of individual tibia. Step 3 retrains STN and AEs to become masked-robust and enable predicting a tibia reconstruction.
• Figure 2: Application of our mask-robust STN+VAE pipeline on three exemplary CTs where fractures have been masked out.