Masked Registration and Autoencoding of CT Images for Predictive Tibia Reconstruction
This work addresses a domain-specific challenge for surgeons in orthopedic surgery, providing an incremental improvement by combining existing neural registration and autoencoder methods with masked input handling.
The paper tackles the problem of predicting patient-specific healthy bone alignment from CT scans of fractured tibias to assist surgical planning, achieving a reconstruction accuracy of 92.5% dice score on a dataset of 150 patients.
Surgical planning for complex tibial fractures can be challenging for surgeons, as the 3D structure of the later desirable bone alignment may be diffi- cult 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 ap- proach 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 varia- tions. 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