Disentangled Latent Representation Learning for Tackling the Confounding M-Bias Problem in Causal Inference

In causal inference, it is a fundamental task to estimate the causal effect from observational data. However, latent confounders pose major challenges in causal inference in observational data, for example, confounding bias and M-bias. Recent data-driven causal effect estimators tackle the confounding bias problem via balanced representation learning, but assume no M-bias in the system, thus they fail to handle the M-bias. In this paper, we identify a challenging and unsolved problem caused by a variable that leads to confounding bias and M-bias simultaneously. To address this problem with co-occurring M-bias and confounding bias, we propose a novel Disentangled Latent Representation learning framework for learning latent representations from proxy variables for unbiased Causal effect Estimation (DLRCE) from observational data. Specifically, DLRCE learns three sets of latent representations from the measured proxy variables to adjust for the confounding bias and M-bias. Extensive experiments on both synthetic and three real-world datasets demonstrate that DLRCE significantly outperforms the state-of-the-art estimators in the case of the presence of both confounding bias and M-bias.

Deep unrolling Shrinkage Network for Dynamic MR imaging

Deep unrolling networks that utilize sparsity priors have achieved great success in dynamic magnetic resonance (MR) imaging. The convolutional neural network (CNN) is usually utilized to extract the transformed domain, and then the soft thresholding (ST) operator is applied to the CNN-transformed data to enforce the sparsity priors. However, the ST operator is usually constrained to be the same across all channels of the CNN-transformed data. In this paper, we propose a novel operator, called soft thresholding with channel attention (AST), that learns the threshold for each channel. In particular, we put forward a novel deep unrolling shrinkage network (DUS-Net) by unrolling the alternating direction method of multipliers (ADMM) for optimizing the transformed $l_1$ norm dynamic MR reconstruction model. Experimental results on an open-access dynamic cine MR dataset demonstrate that the proposed DUS-Net outperforms the state-of-the-art methods. The source code is available at \url{https://github.com/yhao-z/DUS-Net}.

SCULPTOR: Skeleton-Consistent Face Creation Using a Learned Parametric Generator

Recent years have seen growing interest in 3D human faces modelling due to its wide applications in digital human, character generation and animation. Existing approaches overwhelmingly emphasized on modeling the exterior shapes, textures and skin properties of faces, ignoring the inherent correlation between inner skeletal structures and appearance. In this paper, we present SCULPTOR, 3D face creations with Skeleton Consistency Using a Learned Parametric facial generaTOR, aiming to facilitate easy creation of both anatomically correct and visually convincing face models via a hybrid parametric-physical representation. At the core of SCULPTOR is LUCY, the first large-scale shape-skeleton face dataset in collaboration with plastic surgeons. Named after the fossils of one of the oldest known human ancestors, our LUCY dataset contains high-quality Computed Tomography (CT) scans of the complete human head before and after orthognathic surgeries, critical for evaluating surgery results. LUCY consists of 144 scans of 72 subjects (31 male and 41 female) where each subject has two CT scans taken pre- and post-orthognathic operations. Based on our LUCY dataset, we learn a novel skeleton consistent parametric facial generator, SCULPTOR, which can create the unique and nuanced facial features that help define a character and at the same time maintain physiological soundness. Our SCULPTOR jointly models the skull, face geometry and face appearance under a unified data-driven framework, by separating the depiction of a 3D face into shape blend shape, pose blend shape and facial expression blend shape. SCULPTOR preserves both anatomic correctness and visual realism in facial generation tasks compared with existing methods. Finally, we showcase the robustness and effectiveness of SCULPTOR in various fancy applications unseen before.

T$^2$LR-Net: An Unrolling Reconstruction Network Learning Transformed Tensor Low-Rank prior for Dynamic MR Imaging

While the methods exploiting the tensor low-rank prior are booming in high-dimensional data processing and have obtained satisfying performance, their applications in dynamic magnetic resonance (MR) image reconstruction are limited. In this paper, we concentrate on the tensor singular value decomposition (t-SVD), which is based on the Fast Fourier Transform (FFT) and only provides the definite and limited tensor low-rank prior in the FFT domain, heavily reliant upon how closely the data and the FFT domain match up. By generalizing the FFT into an arbitrary unitary transformation of the transformed t-SVD and proposing the transformed tensor nuclear norm (TTNN), we introduce a flexible model based on TTNN with the ability to exploit the tensor low-rank prior of a transformed domain in a larger transformation space and elaborately design an iterative optimization algorithm based on the alternating direction method of multipliers (ADMM), which is further unrolled into a model-based deep unrolling reconstruction network to learn the transformed tensor low-rank prior (T$^2$LR-Net). The convolutional neural network (CNN) is incorporated within the T$^2$LR-Net to learn the best-matched transform from the dynamic MR image dataset. The unrolling reconstruction network also provides a new perspective on the low-rank prior utilization by exploiting the low-rank prior in the CNN-extracted feature domain. Experimental results on two cardiac cine MR datasets demonstrate that the proposed framework can provide improved recovery results compared with the state-of-the-art optimization-based and unrolling network-based methods.

Dynamic MRI using Learned Transform-based Deep Tensor Low-Rank Network (DTLR-Net)

While low-rank matrix prior has been exploited in dynamic MR image reconstruction and has obtained satisfying performance, low-rank tensors models have recently emerged as powerful alternative representations for three-dimensional dynamic MR datasets. In this paper, we introduce a model-based deep learning network by learning the tensor low-rank prior of the cardiac dynamic MR images. Instead of representing the dynamic dataset as a low-rank tensor directly, we propose a learned transformation operator to exploit the tensor low-rank property in a transform domain. In particular, by generalizing the t-SVD tensor decomposition into a unitary transformed t-SVD, we define a transformed tensor nuclear norm (TTNN) to enforce the tensor low-rankness. The dynamic MRI reconstruction problem is thus formulated using a TTNN regularized optimization problem. An iterative algorithm based on ADMM used to minimize the cost is unrolled into a deep network, where the transform is learned using convolutional neural networks (CNNs) to promote the reconstruction quality in the feature domain. Experimental results on cardiac cine MRI reconstruction demonstrate that the proposed framework is able to provide improved recovery results compared with the state-of-the-art algorithms.

Dynamic Cardiac MRI Reconstruction Using Combined Tensor Nuclear Norm and Casorati Matrix Nuclear Norm Regularizations

Low-rank tensor models have been applied in accelerating dynamic magnetic resonance imaging (dMRI). Recently, a new tensor nuclear norm based on t-SVD has been proposed and applied to tensor completion. Inspired by the different properties of the tensor nuclear norm (TNN) and the Casorati matrix nuclear norm (MNN), we introduce a combined TNN and Casorati MNN regularizations framework to reconstruct dMRI, which we term as TMNN. The proposed method simultaneously exploits the spatial structure and the temporal correlation of the dynamic MR data. The optimization problem can be efficiently solved by the alternating direction method of multipliers (ADMM). In order to further improve the computational efficiency, we develop a fast algorithm under the Cartesian sampling scenario. Numerical experiments based on cardiac cine MRI and perfusion MRI data demonstrate the performance improvement over the traditional Casorati nuclear norm regularization method.

Generalized Radiograph Representation Learning via Cross-supervision between Images and Free-text Radiology Reports

Pre-training lays the foundation for recent successes in radiograph analysis supported by deep learning. It learns transferable image representations by conducting large-scale fully-supervised or self-supervised learning on a source domain. However, supervised pre-training requires a complex and labor intensive two-stage human-assisted annotation process while self-supervised learning cannot compete with the supervised paradigm. To tackle these issues, we propose a cross-supervised methodology named REviewing FreE-text Reports for Supervision (REFERS), which acquires free supervision signals from original radiology reports accompanying the radiographs. The proposed approach employs a vision transformer and is designed to learn joint representations from multiple views within every patient study. REFERS outperforms its transfer learning and self-supervised learning counterparts on 4 well-known X-ray datasets under extremely limited supervision. Moreover, REFERS even surpasses methods based on a source domain of radiographs with human-assisted structured labels. Thus REFERS has the potential to replace canonical pre-training methodologies.

nnFormer: Interleaved Transformer for Volumetric Segmentation

Transformers, the default model of choices in natural language processing, have drawn scant attention from the medical imaging community. Given the ability to exploit long-term dependencies, transformers are promising to help atypical convolutional neural networks (convnets) to overcome its inherent shortcomings of spatial inductive bias. However, most of recently proposed transformer-based segmentation approaches simply treated transformers as assisted modules to help encode global context into convolutional representations without investigating how to optimally combine self-attention (i.e., the core of transformers) with convolution. To address this issue, in this paper, we introduce nnFormer (i.e., Not-aNother transFormer), a powerful segmentation model with an interleaved architecture based on empirical combination of self-attention and convolution. In practice, nnFormer learns volumetric representations from 3D local volumes. Compared to the naive voxel-level self-attention implementation, such volume-based operations help to reduce the computational complexity by approximate 98% and 99.5% on Synapse and ACDC datasets, respectively. In comparison to prior-art network configurations, nnFormer achieves tremendous improvements over previous transformer-based methods on two commonly used datasets Synapse and ACDC. For instance, nnFormer outperforms Swin-UNet by over 7 percents on Synapse. Even when compared to nnUNet, currently the best performing fully-convolutional medical segmentation network, nnFormer still provides slightly better performance on Synapse and ACDC.