The Study of Application of Empirical Mode Decomposition to Fractional Anisotropic Diffusion Image for Brain with Schizophrenia

  • 王 儷臻

Student thesis: Master's Thesis

Abstract

Schizophrenia is a severe neuropsychiatric disorder and its hallmark is poor cognitive control In the past schizophrenia was just thought as mental illness but many studies suggested that schizophrenia is intimately related to integrity of white matter in the recently Diffusion tensor image is a non-invasive MR technique which measuring diffusion circumstances of water Fractional anisotropy of DTI represents integrity of white matter The empirical mode decomposition is adaptive and it works well for non-linear and non-stationary signal The mean envelope of intrinsic mode function must be zero at any points in theory but it is difficult to reach in program of EMD Therefore the study adjusts threshold of mean envelope to get better IMFs FA image is decomposed by EMD and statistical analysis of FA image with enhanced IMF is applied to explore that the significant difference of white matter between schizophrenia patients and healthy controls and the relation between hallucination score and FA of specific tracts The prefrontal cortex is related to cognitive control psychosis occurs if their connections go awry The study found that there are significant difference in FA between schizophrenia patients and healthy controls by tract-based spatial and the areas where FA reduce extremely significant are in prefrontal cortex The significantly reduced FA in the superior longitudinal fasciculus and inferior longitudinal fasciculus of patients compared with healthy controls but hallucination score is not related to mean FA of SLF and ILF To sum up the study applied improved EMD on observation of schizophrenia patients’ white matter It found that schizophrenia patients’ connection are significantly poor compared with healthy controls and extremely significantly reduced connection in prefrontal cortex of schizophrenia patients The method is expected to analyze more neuropathy illness
Date of Award2016 Sep 1
Original languageEnglish
SupervisorKuo-Sheng Cheng (Supervisor)

Cite this

'