Weighted Manifold Alignment using Wave Kernel Signatures for Aligning Medical Image Datasets

TY - JOUR

T1 - Weighted Manifold Alignment using Wave Kernel Signatures for Aligning Medical Image Datasets

AU - Clough, James

AU - Balfour, Daniel R.

AU - Cruz, Gastao

AU - Marsden, Paul K.

AU - Prieto , Claudia

AU - Reader, Andrew J.

AU - King, Andrew P.

PY - 2020/4/1

Y1 - 2020/4/1

N2 - Manifold alignment (MA) is a technique to map many high-dimensional datasets to one shared low-dimensional space. Here we develop a pipeline for using MA to reconstruct high-resolution medical images. We present two key contributions. Firstly, we develop a novel MA scheme in which each high-dimensional dataset can be differently weighted preventing noisier or less informative data from corrupting the aligned embedding. We find that this generalisation improves performance in our experiments in both supervised and unsupervised MA problems. Secondly, we use the wave kernel signature as a graph descriptor for the unsupervised MA case finding that it significantly outperforms the current state-of-the-art methods and provides higher quality reconstructed magnetic resonance volumes than existing methods.

AB - Manifold alignment (MA) is a technique to map many high-dimensional datasets to one shared low-dimensional space. Here we develop a pipeline for using MA to reconstruct high-resolution medical images. We present two key contributions. Firstly, we develop a novel MA scheme in which each high-dimensional dataset can be differently weighted preventing noisier or less informative data from corrupting the aligned embedding. We find that this generalisation improves performance in our experiments in both supervised and unsupervised MA problems. Secondly, we use the wave kernel signature as a graph descriptor for the unsupervised MA case finding that it significantly outperforms the current state-of-the-art methods and provides higher quality reconstructed magnetic resonance volumes than existing methods.

U2 - 10.1109/TPAMI.2019.2891600

DO - 10.1109/TPAMI.2019.2891600

M3 - Article

SN - 0162-8828

VL - 42

SP - 988

EP - 997

JO - IEEE Transactions on Pattern Analysis and Machine Intelligence

JF - IEEE Transactions on Pattern Analysis and Machine Intelligence

IS - 4

ER -