Interaction between tumour-infiltrating B cells and T cells controls the progression of hepatocellular carcinoma.

OBJECTIVE: The nature of the tumour-infiltrating leucocytes (TILs) is known to impact clinical outcome in carcinomas, including hepatocellular carcinoma (HCC). However, the role of tumour-infiltrating B cells (TIBs) remains controversial. Here, we investigate the impact of TIBs and their interaction with T cells on HCC patient prognosis. DESIGN: Tissue samples were obtained from 112 patients with HCC from Singapore, Hong Kong and Zurich and analysed using immunohistochemistry and immunofluorescence. RNA expression of CD19, CD8A, IFNG was analysed using quantitative PCR. The phenotype of freshly isolated TILs was analysed using flow cytometry. A mouse model depleted of mature B cells was used for functional study. RESULTS: Tumour-infiltrating T cells and B cells were observed in close contact with each other and their densities are correlated with superior survival in patients with HCC. Furthermore, the density of TIBs was correlated with an enhanced expression of granzyme B and IFN-γ, as well as with reduced tumour viability defined by low expression of Ki-67, and an enhanced expression of activated caspase-3 on tumour cells. CD27 and CD40 costimulatory molecules and TILs expressing activation marker CD38 in the tumour were also correlated with patient survival. Mice depleted of mature B cells and transplanted with murine hepatoma cells showed reduced tumour control and decreased local T cell activation, further indicating the important role of B cells. CONCLUSIONS: The close proximity of tumour-infiltrating T cells and B cells indicates a functional interaction between them that is linked to an enhanced local immune activation and contributes to better prognosis for patients with HCC.

Multiagent off-screen behavior prediction in football.

In multiagent worlds, several decision-making individuals interact while adhering to the dynamics constraints imposed by the environment. These interactions, combined with the potential stochasticity of the agents' dynamic behaviors, make such systems complex and interesting to study from a decision-making perspective. Significant research has been conducted on learning models for forward-direction estimation of agent behaviors, for example, pedestrian predictions used for collision-avoidance in self-driving cars. In many settings, only sporadic observations of agents may be available in a given trajectory sequence. In football, subsets of players may come in and out of view of broadcast video footage, while unobserved players continue to interact off-screen. In this paper, we study the problem of multiagent time-series imputation in the context of human football play, where available past and future observations of subsets of agents are used to estimate missing observations for other agents. Our approach, called the Graph Imputer, uses past and future information in combination with graph networks and variational autoencoders to enable learning of a distribution of imputed trajectories. We demonstrate our approach on multiagent settings involving players that are partially-observable, using the Graph Imputer to predict the behaviors of off-screen players. To quantitatively evaluate the approach, we conduct experiments on football matches with ground truth trajectory data, using a camera module to simulate the off-screen player state estimation setting. We subsequently use our approach for downstream football analytics under partial observability using the well-established framework of pitch control, which traditionally relies on fully observed data. We illustrate that our method outperforms several state-of-the-art approaches, including those hand-crafted for football, across all considered metrics.

Neural scene representation and rendering.

Scene representation-the process of converting visual sensory data into concise descriptions-is a requirement for intelligent behavior. Recent work has shown that neural networks excel at this task when provided with large, labeled datasets. However, removing the reliance on human labeling remains an important open problem. To this end, we introduce the Generative Query Network (GQN), a framework within which machines learn to represent scenes using only their own sensors. The GQN takes as input images of a scene taken from different viewpoints, constructs an internal representation, and uses this representation to predict the appearance of that scene from previously unobserved viewpoints. The GQN demonstrates representation learning without human labels or domain knowledge, paving the way toward machines that autonomously learn to understand the world around them.