Semantic similarity metrics for learned image registration
This work addresses image registration challenges for medical imaging or computer vision applications, but it is incremental as it builds on existing learning-based methods with new metrics.
The paper tackled the problem of image registration by proposing a semantic similarity metric to overcome issues with intensity-based metrics like noise and low contrast, achieving consistently high registration accuracy across multiple modalities and smoother transformations on low-quality images.
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.