Li, Chunyuan | Engineering Papers

DOE OSTI2019

EMT-related protein expression in polyploid giant cancer cells and their daughter cells with different passages after triptolide treatment

Our previous work has demonstrated that paclitaxel can induce the formation of polyploid giant cancer cells (PGCCs) and inhibit tumor growth by reprogramming ovarian cancer epithelial cells to a benign fibroblastic state via epithelial–mesenchymal transition. Here, triptolide (TPL) was used to treat the breast and ovarian cancer lines. The morphologic characteristics and EMT-related protein expression were studied in different generation of cancer cells after TPL treatment. When BT-549 and HEY cells reached 80–90% confluence, TPL was added to BT-549 for 48 h and HEY for 9 h at a concentration of 40 ng/ml. Scattered PGCCs survived from TPL treatment and generated daughter cells, and then were cultured in medium without TPL for at least ten generation. Western blot analysis and immunocytochemical staining were performed to detect the expression levels and subcellular location of EMT-related proteins in control cells and different generation of TPL-induced PGCCs with daughter cells. Furthermore, wound-healing, transwell, cell counting kit-8, and MTT assay were used to compare the alternation of migration, invasion, and proliferation among control cells and different generation of TPL-induced PGCCs with daughter cells. Scattered PGCCs survived from the treatment of TPL and produced small-sized daughter cells 20–30 days after treatment. Compared to the control cells, the first generation of TPL-induced PGCCs with their daughter cells differentially expressed EMT-related proteins including fibronectin, E-cadherin, vimentin, and Twist, and had lower migration, invasion, and proliferation abilities. The abilities of migration, invasion, and proliferation of TPL-induced PGCCs with their daughter cells gradually enhanced as the passages increasing, and markedly exceeded the control cells in the tenth generation. TPL-induced PGCCs with their daughter cells gradually obtain the abilities of invasion and metastasis in vitro as the number of passage increasing, which can be used to mimick the cancer cells subjected to anti-cancer drugs in vivo and may provide some new insights to explore the mechanism of cancer invasion, metastasis and relapse after chemotherapy.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

DOE OSTI2016

Learning Weight Uncertainty with Stochastic Gradient MCMC for Shape Classification

Learning the representation of shape cues in 2D & 3D objects for recognition is a fundamental task in computer vision. Deep neural networks (DNNs) have shown promising performance on this task. Due to the large variability of shapes, accurate recognition relies on good estimates of model uncertainty, ignored in traditional training of DNNs, typically learned via stochastic optimization. This paper leverages recent advances in stochastic gradient Markov Chain Monte Carlo (SG-MCMC) to learn weight uncertainty in DNNs. It yields principled Bayesian interpretations for the commonly used Dropout/DropConnect techniques and incorporates them into the SG-MCMC framework. Extensive experiments on 2D & 3D shape datasets and various DNN models demonstrate the superiority of the proposed approach over stochastic optimization. Our approach yields higher recognition accuracy when used in conjunction with Dropout and Batch-Normalization.

convolutional neural network↗

DOE OSTI2016

A Deep Generative Deconvolutional Image Model

A deep generative model is developed for representation and analysis of images, based on a hierarchical convolutional dictionary-learning framework. Stochastic unpooling is employed to link consecutive layers in the model, yielding top-down image generation. A Bayesian support vector machine is linked to the top-layer features, yielding max-margin discrimination. Deep deconvolutional inference is employed when testing, to infer the latent features, and the top-layer features are connected with the max-margin classifier for discrimination tasks. The model is efficiently trained using a Monte Carlo expectation-maximization (MCEM) algorithm; the algorithm is implemented on graphical processor units (GPU) to enable large-scale learning, and fast testing. Excellent results are obtained on several benchmark datasets, including ImageNet, demonstrating that the proposed model achieves results that are highly competitive with similarly sized convolutional neural networks.

convolutional neural network↗