%0 Journal Article
%A Wulan, Naren
%A An, Lijun
%A Zhang, Chen
%A Kong, Ru
%A Chen, Pansheng
%A Bzdok, Danilo
%A Eickhoff, Simon B.
%A Holmes, Avram J.
%A Yeo, B. T. Thomas
%T Translating phenotypic prediction models from big to small anatomical MRI data using meta-matching
%J Imaging neuroscience
%V 2
%@ 2837-6056
%C Cambridge, MA
%I MIT Press
%M FZJ-2024-05517
%P 1 - 21
%D 2024
%X Individualized phenotypic prediction based on structural magnetic resonance imaging (MRI) is an important goal in neuroscience. Prediction performance increases with larger samples, but small-scale datasets with fewer than 200 participants are often unavoidable. We have previously proposed a “meta-matching” framework to translate models trained from large datasets to improve the prediction of new unseen phenotypes in small collection efforts. Meta-matching exploits correlations between phenotypes, yielding large improvement over classical machine learning when applied to prediction models using resting-state functional connectivity as input features. Here, we adapt the two best performing meta-matching variants (“meta-matching finetune” and “meta-matching stacking”) from our previous study to work with T1-weighted MRI data by changing the base neural network architecture to a 3D convolution neural network. We compare the two meta-matching variants with elastic net and classical transfer learning using the UK Biobank (N = 36,461), the Human Connectome Project Young Adults (HCP-YA) dataset (N = 1,017), and the HCP-Aging dataset (N = 656). We find that meta-matching outperforms elastic net and classical transfer learning by a large margin, both when translating models within the same dataset and when translating models across datasets with different MRI scanners, acquisition protocols, and demographics. For example, when translating a UK Biobank model to 100 HCP-YA participants, meta-matching finetune yielded a 136% improvement in variance explained over transfer learning, with an average absolute gain of 2.6% (minimum = –0.9%, maximum = 17.6%) across 35 phenotypes. Overall, our results highlight the versatility of the meta-matching framework.
%F PUB:(DE-HGF)16
%9 Journal Article
%$ 40800257
%U <Go to ISI:>//WOS:001531565300003
%R 10.1162/imag_a_00251
%U https://juser.fz-juelich.de/record/1030924