… the correlations among various modality-specific information and … Disentanglement Network
(D2-Net) for brain tumor segmentation with missing modalities, which consists of a modality …

… networks, but we apply it across all modalities obviating the need for a combinatorial number
of inference networks. In … interactions but without considering missing or multimodal data. …

… use the existing modality to impute the missing modality and may … disentanglement module
to split the information of existing modality into two parts, and then perform missing modality …

… across any subset of modalities and therefore handles missing modalities efficiently.
Disentanglement. Our goal is not the unsupervised disentanglement of all generative factors, which …

… ; 4) we proposed an imputation algorithm for missing modality data by employing the
shared information and specific information represented by the disentangled latent variable. …

… 2, we show that with the increase of the number of missing modalities, the segmentation
results produced by our robust model just gradually degrade, rather than encountering sudden …

… the disentangled anatomical representations as a set of modality-… disentangled
representations in combination with the fusion strategy (Sect. 3.2) can alleviate the missing modality …

… may have incomplete modalities. For … missing modality in terms of flexibility (missing
modalities in training, testing, or both) and efficiency (most training data have incomplete modality). …

… , the single-level generation is not able to impute the missing modality as we have no
context information to recover their semantic trajectories. Instead, by using the inter-…

… Since the model assumes full modality, we compensate for the missing modality CXR with
all zeros during training and testing. DAFT (Pölsterl, Wolf, and Wachinger 2021) is a module …