Integrated Calendar

The cortical system for processing faces is a model system for studying the functional neuroanatomy of ventral temporal cortex and its role in perception for two reasons. First, the functional organization of the cortical face system is well understood. Second, activations in ventral face-selective regions are causally related to face perception. Here, I will describe recent results from our research elucidating the computations performed by population receptive field (pRFs) in the cortical system for face perception. In contrast to predictions of classical theories, recent data from my lab reveals that computations in face-selective regions in human ventral temporal cortex can be characterized with a computational pRF model, which predicts the location and spatial extent of the visual field that is processed by the neural population in a voxel. Our research characterizes pRF properties of ventral face-selective regions revealing three main findings. First, pRFs illustrate a hierarchical organization within the face system, whereby pRFs become larger and more foveal across the ventral hierarchy. Second, attention to faces modulates pRFs in face-selective regions, consequently enhancing the representation of faces in the peripheral visual field where visual acuity is the lowest. Third, our research shows that pRF properties in face-selective regions are behaviorally relevant. We find that face perception abilities are correlated with pRF properties: participants with larger pRFs perform better in face recognition than participants with smaller pRFs. These data suggest that computations performed by pRFs in face-selective regions may form a neural basis for holistic processing necessary for face recognition. Overall, these data highlight the importance of elucidating computational properties of neural populations in ventral temporal cortex as they offer a new mechanistic understanding of high-level visual processes such as face perception.

In vitro model systems used in drug discovery typically do not address the complex architecture of human disease tissues. Only few approaches aim to faithfully recapitulate the complexity, heterogeneity and cellular dynamics e.g. in epithelial tissues and carcinomas. The most important aspects relate to the (tumor-) microenvironment, including cell-cell and cell-matrix interactions, inflammation and the role of stromal components. All of these elements can have a significant, but often underestimated impact on differentiation, normal and abnormal tissue functions, or drug response versus drug resistance.

The basis for performing informative high content screening campaigns with such complex tissue models in vitro is access to fast, automated image analysis. We have developed a software platform (AMIDA, Automated Morphometric Image Data Analysis) that captures a large number of morphometric features in an unsupervised fashion. This approach enables us to capture much of the inherent complexity and dynamics of microtissues, yet still allows high experimental throughput. This screening platform is ideally suited for investigating a broad spectrum of defined, biological questions in drug discovery as well as personalised medicine.

Technology and screening platform are applicable for multiple types of research, such as quantitatively measuring the response of primary cancer cells or cell lines to drugs, siRNAs or other perturbations. Image analysis algorithms can also be adapted towards specific applications in neurodegenerative diseases, stem cell research, and to quantitate the interaction of epithelial cells with immune, adipocytes or mesenchymal stem cells.