CoverBLIP: scalable iterative matched filtering for MR Fingerprint recovery

TL;DR

CoverBLIP introduces an iterative matched filtering approach for MR Fingerprint recovery that enhances accuracy and scalability by leveraging Cover trees and approximate nearest neighbor algorithms, overcoming limitations of traditional linear compression methods.

Contribution

The paper presents CoverBLIP, a novel iterative reconstruction method that improves accuracy and scalability in MR Fingerprint recovery by integrating Cover trees with approximate nearest neighbor search.

Findings

Enhanced accuracy over non-iterative methods

Improved scalability with Cover trees

Robustness against high-dimensional challenges

Abstract

Current proposed solutions for the high dimensionality of the MRF reconstruction problem rely on a linear compression step to reduce the matching computations and boost the efficiency of fast but non-scalable searching schemes such as the KD-trees. However such methodologies often introduce an unfavourable compromise in the estimation accuracy when applied to nonlinear data structures such as the manifold of Bloch responses with possible increased dynamic complexity and growth in data population. To address this shortcoming we propose an inexact iterative reconstruction method, dubbed as the Cover BLoch response Iterative Projection (CoverBLIP). Iterative methods improve the accuracy of their non-iterative counterparts and are additionally robust against certain accelerated approximate updates, without compromising their final accuracy. Leveraging on these results, we accelerate matched-filtering using an ANNS algorithm based on Cover trees with a robustness feature against the curse of dimensionality.