i3PosNet: Instrument Pose Estimation from X-Ray in temporal bone surgery

TL;DR

i3PosNet is a deep learning framework that accurately estimates surgical instrument pose from X-ray images, outperforming traditional methods and generalizing from synthetic to real data in minimally invasive temporal bone surgery.

Contribution

The paper introduces i3PosNet, a novel deep learning approach for instrument pose estimation that works with synthetic training data and generalizes to real X-ray images.

Findings

Achieves errors less than 0.05mm in pose estimation.

Reduces average and maximum errors by at least two thirds compared to traditional methods.

Generalizes from synthetic to real X-ray images without additional training.

Abstract

Purpose: Accurate estimation of the position and orientation (pose) of surgical instruments is crucial for delicate minimally invasive temporal bone surgery. Current techniques lack in accuracy and/or line-of-sight constraints (conventional tracking systems) or expose the patient to prohibitive ionizing radiation (intra-operative CT). A possible solution is to capture the instrument with a c-arm at irregular intervals and recover the pose from the image. Methods: i3PosNet infers the position and orientation of instruments from images using a pose estimation network. Said framework considers localized patches and outputs pseudo-landmarks. The pose is reconstructed from pseudo-landmarks by geometric considerations. Results: We show i3PosNet reaches errors less than 0.05mm. It outperforms conventional image registration-based approaches reducing average and maximum errors by at least two thirds. i3PosNet trained on synthetic images generalizes to real x-rays without any further adaptation. Conclusion: The translation of Deep Learning based methods to surgical applications is difficult, because large representative datasets for training and testing are not available. This work empirically shows sub-millimeter pose estimation trained solely based on synthetic training data.