Publications

Conference
#

2024

✨BMFT: Achieving Fairness via Bias-based Weight Masking Fine-tuning
MICCAI
Oct, 2024
Yuyang Xue, J Yan, R Dutt, F Haider, J Liu, S McDonagh, SA Tsaftaris
Abstract: we propose Bias-based Weight Masking Fine-Tuning (BMFT), a novel post-processing method that enhances the fairness of a trained model in significantly fewer epochs without requiring access to the original training data. BMFT produces a mask over model parameters, which efficiently identifies the weights contributing the most towards biased predictions.
[Paper][Code]
✨Erase to Enhance: Data-Efficient Machine Unlearning in MRI Reconstruction
MIDL
Feb, 2024
Yuyang Xue, Jingshuai Liu, Steven McDonagh, Sotirios A Tsaftaris
Abstract: Our study reveals that combin- ing training data can lead to hallucinations and reduced image quality in the reconstructed data. We use unlearning to remove hallucinations as a proxy exemplar of undesired data removal. Indeed, we show that machine unlearning is possible without full retraining. Furthermore, our observations indicate that maintaining high performance is feasible even when using only a subset of retain data.
[Paper][Code]
✨Inference Stage Denoising for Undersampled MRI Reconstruction
ISBI
Feb, 2024
Yuyang Xue, Chen Qin, Sotirios A Tsaftaris
Abstract: In this work, by employing a conditional hyperparameter network, we eliminate the need of augmentation, yet maintain robust performance under various levels of Gaussian noise. We demonstrate that our model withstands various input noise levels while producing high-definition reconstructions during the test stage. Moreover, we present a hyperparameter sampling strategy that accelerates the convergence of training.

[Paper][Code]
✨Cine cardiac MRI reconstruction using a convolutional recurrent network with refinement
CMRxRecon
Feb, 2024
Yuyang Xue, Y Du, G Carloni, E Pachetti, C Jordan, SA Tsaftaris
Abstract: In this challenge paper, we investigate the use of a convolutional recurrent neural network (CRNN) architecture to exploit temporal correlations in supervised cine cardiac MRI reconstruction. This is combined with a single-image super-resolution refinement module to improve single coil reconstruction by 4.4% in structural similarity and 3.9% in normalised mean square error compared to a plain CRNN implementation.

[Paper][Code]