Consistent Point Matching

TL;DR

This paper introduces a consistency heuristic for point-matching in medical images, enhancing robustness and accuracy across modalities without machine learning, validated on multiple datasets including Deep Lesion Tracking.

Contribution

It presents a novel consistency-based point-matching algorithm that improves robustness and accuracy in medical image registration without requiring training data.

Findings

Outperforms state-of-the-art on Deep Lesion Tracking

Effective in landmark localization tasks

Operates efficiently on standard CPU hardware

Abstract

This study demonstrates that incorporating a consistency heuristic into the point-matching algorithm \cite{yerebakan2023hierarchical} improves robustness in matching anatomical locations across pairs of medical images. We validated our approach on diverse longitudinal internal and public datasets spanning CT and MRI modalities. Notably, it surpasses state-of-the-art results on the Deep Lesion Tracking dataset. Additionally, we show that the method effectively addresses landmark localization. The algorithm operates efficiently on standard CPU hardware and allows configurable trade-offs between speed and robustness. The method enables high-precision navigation between medical images without requiring a machine learning model or training data.