Scalable Query Answering Under Uncertainty to Neuroscientific Ontological Knowledge: The NeuroLang Approach.

Researchers in neuroscience have a growing number of datasets available to study the brain, which is made possible by recent technological advances. Given the extent to which the brain has been studied, there is also available ontological knowledge encoding the current state of the art regarding its different areas, activation patterns, keywords associated with studies, etc. Furthermore, there is inherent uncertainty associated with brain scans arising from the mapping between voxels-3D pixels-and actual points in different individual brains. Unfortunately, there is currently no unifying framework for accessing such collections of rich heterogeneous data under uncertainty, making it necessary for researchers to rely on ad hoc tools. In particular, one major weakness of current tools that attempt to address this task is that only very limited propositional query languages have been developed. In this paper we present NeuroLang, a probabilistic language based on first-order logic with existential rules, probabilistic uncertainty, ontologies integration under the open world assumption, and built-in mechanisms to guarantee tractable query answering over very large datasets. NeuroLang's primary objective is to provide a unified framework to seamlessly integrate heterogeneous data, such as ontologies, and map fine-grained cognitive domains to brain regions through a set of formal criteria, promoting shareable and highly reproducible research. After presenting the language and its general query answering architecture, we discuss real-world use cases showing how NeuroLang can be applied to practical scenarios.

Neural networks and artificial intelligence in thoracic surgery.

The human brain has billions of neurons and connections that cannot be emulated by computers. This structure could explain the anatomical basis of typically human psychological activities like intuition or artistic creation. On the other hand, the computer-organized way of "reasoning" binary problems through systematic comparison of a large number of data-as AIM does--is impossible to be emulated by humans. At the same time, AIM, through the use of different methods like ANN or DM systems, is able to give individualized answers to otherwise probabilistic population problems. Hence, that is the reason for its application in the assessment of surgical risk in lung resection candidates. With regard to AIM methodology, many issues could be addressed and argued, especially on the data collection because of the retrospective nature of the data on which the available contributions from the literature are based. In the larger studies, patients from different centers treated by different surgical teams were included. Both circumstances could have caused heterogeneity of the study groups, which, in turn, can lead to less-reliable conclusions. Even if limited, our experience became an appealing one because AIM seems to be a potentially useful complementary tool to the nonreplaceable clinical judgment.