Collaborative Quantization Embeddings for Intra-Subject Prostate MR Image Registration

TL;DR

This paper introduces a hierarchical quantization and collaborative dictionary approach to improve learning-based intra-subject prostate MRI registration, enhancing accuracy and generalization with limited clinical data.

Contribution

It proposes a novel hierarchical quantization method and a collaborative dictionary to incorporate prior information, improving registration accuracy and robustness in prostate MRI analysis.

Findings

Significant improvement in registration accuracy with 28.7% error reduction.

Enhanced generalization between training and testing data.

Effective incorporation of prior information like gland segmentation.

Abstract

Image registration is useful for quantifying morphological changes in longitudinal MR images from prostate cancer patients. This paper describes a development in improving the learning-based registration algorithms, for this challenging clinical application often with highly variable yet limited training data. First, we report that the latent space can be clustered into a much lower dimensional space than that commonly found as bottleneck features at the deep layer of a trained registration network. Based on this observation, we propose a hierarchical quantization method, discretizing the learned feature vectors using a jointly-trained dictionary with a constrained size, in order to improve the generalisation of the registration networks. Furthermore, a novel collaborative dictionary is independently optimised to incorporate additional prior information, such as the segmentation of the gland or other regions of interest, in the latent quantized space. Based on 216 real clinical images from 86 prostate cancer patients, we show the efficacy of both the designed components. Improved registration accuracy was obtained with statistical significance, in terms of both Dice on gland and target registration error on corresponding landmarks, the latter of which achieved 5.46 mm, an improvement of 28.7\% from the baseline without quantization. Experimental results also show that the difference in performance was indeed minimised between training and testing data.