Computer aided diagnosis of functional brain disorders using PET on a PC based platform

We describe a portable PC based platform for computer aided diagnosis (CAD) of functional brain disorders using positron emission tomography (PET) that integrates a variety of methods including quantitative and visual techniques as well as extensively automated procedures. Statistical procedures such as the scaled subprofile model (SSM) in conjunction with principal component analysis are employed to derive neuronal networks that can uniquely identify characteristic metabolic covariance patterns. By measuring the expression of these networks in diseased subjects on an individual basis we can help automate the diagnosis of conditions such as Parkinson's disease and evaluate treatment efficacy and disease progression.