Blood vessel feature description for detection of Alzheimers disease

摘要

We describe how image analysis can be used to detect the presence of Alzheimer's disease. The data are images of brain tissue collected from subjects with and without Alzheimer's disease. The analysis concentrates on the shape and structure of the blood vessels which are known to be affected by amyloid beta, whose drainage is affected by Alzheimer's disease. The structure is analysed by a new approach which measures the Influence of the blood vessels' branching structures. Their density and tortuosity are analysed in conjunction with a boundary description derived using Fourier descriptors. These measures form a feature vector which is derived from the images of brain tissue, and the discrimination capability shows that it is possible to detect the presence of Alzheimer's disease using these measures and in an automated way. These measures also show that shape information is influenced by the vessels' branching structure, as known to be consistent with Alzheimer's disease evolution.