Studies over the past decade have generated a wealth of molecular data that can be leveraged to better understand cancer risk, progression, and outcomes. However, understanding the progression risk and differentiating long- and short-term survivors cannot be achieved by analyzing data from a single modality due to the heterogeneity of disease. Using a scientifically developed and tested deep-learning approach that leverages aggregate information collected from multiple repositories with multiple modalities (e.g., mRNA, DNA Methylation, miRNA) could lead to a more accurate and robust prediction of disease progression. Here, we propose an autoencoder based multimodal data fusion system, in which a fusion encoder flexibly integrates collective information available through multiple studies with partially coupled data. Our results on a fully controlled simulation-based study have shown that inferring the missing data through the proposed data fusion pipeline allows a predictor that is superior to other baseline predictors with missing modalities. Results have further shown that short- and long-term survivors of glioblastoma multiforme, acute myeloid leukemia, and pancreatic adenocarcinoma can be successfully differentiated with an AUC of 0.94, 0.75, and 0.96, respectively.

Recent supervised point cloud upsampling methods are restricted by the size of training data and are limited in terms of covering all object shapes. Besides the challenges faced due to data acquisition, the networks also struggle to generalize on unseen records. In this paper, we present an internal point cloud upsampling approach at a holistic level referred to as "Zero-Shot" Point Cloud Upsampling (ZSPU). Our approach is data agnostic and relies solely on the internal information provided by a particular point cloud without patching in both self-training and testing phases. This single-stream design significantly reduces the training time by learning the relation between low resolution (LR) point clouds and their high (original) resolution (HR) counterparts. This association will then provide super resolution (SR) outputs when original point clouds are loaded as input. ZSPU achieves competitive/superior quantitative and qualitative performances on benchmark datasets when compared with other upsampling methods.

Numerous point-cloud understanding techniques focus on whole entities and have succeeded in obtaining satisfactory results and limited sparsity tolerance. However, these methods are generally sensitive to incomplete point clouds that are scanned with flaws or large gaps. In this paper, we propose an end-to-end architecture that compensates for and identifies partial point clouds on the fly. First, we propose a cascaded solution that integrates both the upstream masked autoencoder (MAE) and downstream understanding networks simultaneously, allowing the task-oriented downstream to identify the points generated by the completion-oriented upstream. These two streams complement each other, resulting in improved performance for both completion and downstream-dependent tasks. Second, to explicitly understand the predicted points’ pattern, we introduce hierarchical self-distillation (HSD), which can be applied to any hierarchy-based point cloud methods. HSD ensures that the deepest classifier with a larger perceptual field of local kernels and longer code length provides additional regularization to intermediate ones rather than simply aggregating the multi-scale features, and therefore maximizing the mutual information (MI) between a teacher and students. The proposed HSD strategy is particularly well-suited for tasks involving scattered point clouds, wherein a singular prediction may yield imprecise outcomes due to the inherently irregular and sparse nature of the geometric shape being reconstructed. We show the advantage of the self-distillation process in the hyperspaces based on the information bottleneck principle. Our method achieves state-of-the-art on both classification and part segmentation tasks.