Multi-task Model Based On Cortical Morphology Of The Cortex And Its Classification Research

The brain is the most important and complex organ in the human body,the equivalent of the computer CPU,dominate the people’s thinking and various acts,study the brain’s organizational structure and working mechanism for human health is of great significance.Alzheimer’s disease(AD)is a degenerative disease of the nervous system,and the onset and development of the disease is not easily found by people,clinical memory,communication and language skills,behavioral competence,cognitive function Sustained decline or even total loss,leading to a comprehensive dementia,folk called dementia.Mild cognitive impairment(MCI)is an intermediate state between normal aging and dementia.Compared with normal control(NC)elderly patients,patients with mild cognitive impairment,but not enough to show dementia,Is the high incidence of senile dementia.Due to the current can not cure Alzheimer’s disease,the condition is irreversible,has not been converted into Alzheimer’s mild cognitive impairment of this stage of prevention and intervention,for the control of patients with disease and delayed dementia Have a significant effect.In a growing number of related studies have shown that AD patients and normal elderly,MCI patients and normal elderly brain tissue structure has changed,there is a difference between the cortex,with magnetic resonance imaging(MRI)technology,It is important to study the changes of AD and MCI brain structure and to reduce the morbidity and mortality of AD,and to provide theoretical basis for early prediction and automatic diagnosis of AD.The main contributions of this paper are as follows:1.Three kinds of brain morphological indexes were calculated based on the cerebral cortex morphological analysis of AD brain samples of AD,MCI and NC subjects.Cortex thickness,gray matter volume and cortical complexity were calculated.Based on each index data,the difference of brain structure between each two groups was compared with each index data.It was proved that there was significant difference between brain tissue of AD and NC and MCI and NC model in each index,To provide the basis for multi-task learning to select the characteristics,while finding out what brain area abnormalities.2.The AD and NC,MCI and NC models were respectively classified.In this field,we try to use the multi-task learning(MTL)method to select the characteristics of the three indicators of cerebral cortical morphology: cortex thickness,gray matter volume and cortical complexity as three tasks,select the characteristics of the task between the existence of a certain The contact.At the same time,it is compared with F-score and mRMR.On the basis of feature selection,it is more important to select a suitable classifier.In this paper,we try to use the classification method of extreme learning machine(ELM),and combine SVM-Linear and SVM-RBF to explore the commonly used The optimal combination of feature selection method and classifier in AD and NC and MCI and NC classification.