Semi-Automatic Framework for Estimating Intelligence Quotient (IQ) Using MRI

Abstract

Central Nervous System is the huge database of information, called a well-organized intellectual library. Information storage, retrieval and transfer in neurons is enormouslyfast and feels like wide and advanced network placed inside our skull. From literature survey we conclude that only some regions of the brain is specifically contributing to intelligence. The central regions are identified and measured in various dimensions for statistical analysis. For analyzing IQ score a popular imaging called MRI(Magnetic Resonance Imaging)is used. The positively associated regions related to intelligence are total brain volume, GM/WM and cortical analysis. This analysis mainly focused in consolidating all the brain variables that are correlated with intelligence. Identifying and measuring only the few particular variables cannot determine intelligence, this motivated us in considering all the variables at once. Brain Morphometric features are measured manually or automatic using volumetric analysis of core regions and sub regions. The some of the morphometric analysis are volumetric analysis and quantification of size and shape of the brain’s variables. Here in this proposed framework we used  subject’s MRI in estimating human’s intelligence using MeVisLab and Brainsuite19b. We also presented various graphical network framework using itk and vtk packages of MeVisLab. Our results consolidated the vital regions and considered its overall volumetric measures in improving the intelligence accuracy.Future advancement of this analysis is to define particularly that which region is specifically associated with what intelligence types.  To check whether the individual  will excel in particular field the corresponding associated region and its measurementswillbe useful in predicting the intelligence level and also stating whether this subject can be suitable or not for the particular field of interest.