Binary and nonbinary description of hypointensity in human brain MR images

TL;DR

This paper introduces improved binary and novel nonbinary methods for describing hypointensity in brain MR images to better understand iron accumulation related to neurodegenerative diseases.

Contribution

It proposes new binary and nonbinary hypointensity descriptions based on principal components analysis and compares their information content using Kendall's rank correlation.

Findings

Enhanced binary hypointensity description method

Novel nonbinary hypointensity description using PCA

Complementary insights into iron accumulation in brain tissues

Abstract

Accumulating evidence has shown that iron is involved in the mechanism underlying many neurodegenerative diseases, such as Alzheimer's disease, Parkinson's disease and Huntington's disease. Abnormal (higher) iron accumulation has been detected in the brains of most neurodegenerative patients, especially in the basal ganglia region. Presence of iron leads to changes in MR signal in both magnitude and phase. Accordingly, tissues with high iron concentration appear hypo-intense (darker than usual) in MR contrasts. In this report, we proposed an improved binary hypointensity description and a novel nonbinary hypointensity description based on principle components analysis. Moreover, Kendall's rank correlation coefficient was used to compare the complementary and redundant information provided by the two methods in order to better understand the individual descriptions of iron accumulation in the brain.