Semantic similarity metrics for learned image registration

TL;DR

This paper introduces a semantic similarity metric for image registration that leverages learned features to improve accuracy and robustness across various modalities and noise conditions.

Contribution

It presents a novel learned semantic similarity metric using unsupervised and semi-supervised methods, enhancing registration performance over traditional intensity-based metrics.

Findings

Achieves higher registration accuracy across multiple modalities.

Provides smoother transformations on low-quality images.

Demonstrates robustness to noise through learned invariance.

Abstract

We propose a semantic similarity metric for image registration. Existing metrics like Euclidean Distance or Normalized Cross-Correlation focus on aligning intensity values, giving difficulties with low intensity contrast or noise. Our approach learns dataset-specific features that drive the optimization of a learning-based registration model. We train both an unsupervised approach using an auto-encoder, and a semi-supervised approach using supplemental segmentation data to extract semantic features for image registration. Comparing to existing methods across multiple image modalities and applications, we achieve consistently high registration accuracy. A learned invariance to noise gives smoother transformations on low-quality images.