Database system for patient-based neuroscience research

University Of Pennsylvania, Philadelphia PA

Investigators

Linked publications & trials

Abstract

Study of brain-injured patients has long been a critical method in understanding cognitive function in humans. However, methods and tools for managing data for patient-based studies - including demographic, behavioral, image, and lesion location data - have lagged behind. Such tools are especially critical to patient-based research given that finding patients who meet the inclusion/exclusion requirements of a given study is often a bottleneck in completing studies. Ideally, such tools should facilitate both local and collaborative research while conforming to regulatory constraints. However, at present many researchers rely on ad-hoc data management approaches that don't support the full range of databasing needs. We propose to develop a free software system for managing patient data, appropriate to the needs of researchers studying cognitive function in neurological populations. Our proposal includes two main components. The first is a portable, free database system designed to be easily adapted to local needs and to facilitate sharing resources to the extent permitted by prevailing regulations. The system is extensible, readily installed on a variety of platforms, and suitable for a range of scales, from desktop use for individual researchers to inter-institutional collaborations. The second component of our proposal is a system for managing and combining lesion location information of widely varying quality, ranging from high-resolution MRI to hand-drawn sketches, by incorporating localization error into the assessment of the probability that a lesion encompasses or overlaps a given region. This system will allow the construction of probabilistic lesion maps for use in research, as well as database search on the basis of lesion location, with probabilistic rather than all-or-none criteria.