Brain tissues volume measurements from 2D MRI using parametric approach
Alexey Arlenovich L’vov, Yu. V. Litovka, Olga Toropova · 2018
The purpose of the paper is to propose a fully automated method of volume assessment of structures within human brain. Our statistical approach uses maximum interdependency principle for decision making process of measurements consistency and unequal observations. Detecting outliers performed using maximum normalized residual test. We propose a statistical model which utilizes knowledge of tissues distribution in human brain and applies partial data restoration for precision improvement. The approach proposes completed computationally efficient and independent from segmentation algorithm used in the application.