Brain image feature recognition method for Alzheimer’s disease

Abstract

To study the early automatic recognition of Alzheimer’s disease, Alzheimer’s disease automatic recognition method is explored based on brain magnetic resonance of brain asymmetry image features. The method, according to the atrophy of related brain tissue of patients with Alzheimer’s disease, leading to the pathological characteristics of the asymmetry of left and right brain related anatomical structures, proposed to extract the shape and texture features of these anatomical structure as the asymmetry, and selected out the best feature subset that can characterize the lesion index as the method for automatic recognition of early Alzheimer’s disease. The automatic recognition test of the patient images in Southwest Hospital was performed by this method and compared with the expert automatic recognition. The results show that the automatic recognition method is obviously effective. Based on the above findings, it is concluded that the automatic recognition method of Alzheimer’s disease has good performance.