本文選題：糖尿病 + 遺忘型輕度認(rèn)知障礙��； 參考：《燕山大學(xué)》2015年碩士論文

【摘要】：糖尿病及其引起認(rèn)知功能損害的患病率正在逐年上升,嚴(yán)重影響到了患者的生活質(zhì)量。糖尿病是導(dǎo)致認(rèn)知功能下降和發(fā)展成癡呆的一個(gè)重要危險(xiǎn)因素。研究糖尿病與認(rèn)知功能相關(guān)的腦電信號(hào),可加深人們對(duì)糖尿病認(rèn)知功能損害機(jī)制的了解,對(duì)早期干預(yù)有重大意義。首先,研究經(jīng)驗(yàn)?zāi)B(tài)分解的近似熵、樣本熵、模糊熵、排序熵、功率譜熵和小波熵,利用Logistic映射模型進(jìn)行仿真,仿真結(jié)果表明模糊熵的抗噪性較好。分析糖尿病患者遺忘型輕度認(rèn)知障礙組和對(duì)照組腦電信號(hào)的非線性變化,采用單因素方差統(tǒng)計(jì)分析,基于曲線下最大面積的特征提取和支持向量機(jī)進(jìn)行分類,對(duì)熵與神經(jīng)心理學(xué)測(cè)試進(jìn)行皮爾森線性相關(guān)分析。結(jié)果表明模糊熵的分類效果更好,前額、右顳和枕區(qū)為顯著腦區(qū),熵與神經(jīng)心理學(xué)測(cè)試在枕區(qū)和顳區(qū)存在相關(guān)性。其次,將符號(hào)化模式與遞歸圖相結(jié)合。符號(hào)化模式遞歸圖對(duì)采樣量及平穩(wěn)性要求較低,受噪聲影響小,可以直接對(duì)時(shí)間序列進(jìn)行編碼。利用Logistic映射和單通道雙動(dòng)態(tài)神經(jīng)元群模型進(jìn)行仿真,結(jié)果表明符號(hào)化模式遞歸圖的確定性變量比排序遞歸圖可以更好地刻畫模型參數(shù)的變化。最后,對(duì)實(shí)際糖尿病和非糖尿病中的遺忘型輕度認(rèn)知障礙組和對(duì)照組分析,結(jié)果表明符號(hào)化模式遞歸圖比排序遞歸圖的確定性變量能夠更好地區(qū)別兩組。在認(rèn)知狀態(tài)匹配下,發(fā)現(xiàn)糖尿病患者與認(rèn)知功能存在某種聯(lián)系。確定性變量與神經(jīng)心理學(xué)測(cè)試在前額、顳區(qū)和枕區(qū)存在相關(guān)性。
[Abstract]:The prevalence of diabetes mellitus and its cognitive impairment is increasing year by year, which seriously affects the quality of life of patients. Diabetes is an important risk factor for cognitive impairment and the development of dementia. The study of the EEG associated with diabetes mellitus and cognitive function can deepen the understanding of the mechanism of diabetes cognitive impairment and have great significance for early intervention. Firstly, the approximate entropy, sample entropy, fuzzy entropy, sort entropy, power spectrum entropy and wavelet entropy of empirical mode decomposition are studied. The simulation results show that the fuzzy entropy has good noise resistance. The nonlinear changes of EEG in amnesia group and control group were analyzed. Single factor variance statistical analysis was used to extract the largest area of feature under the curve and classify it with support vector machine (SVM). Pearson linear correlation analysis of entropy and neuropsychological tests was carried out. The results showed that the classification effect of fuzzy entropy was better. The frontal, right temporal and occipital regions were significant brain regions, and the correlation between entropy and neuropsychological tests was found in occipital and temporal regions. Secondly, the symbolic pattern is combined with the recursive graph. The symbolic pattern recursion graph requires less sampling amount and stability, and is less affected by noise, so the time series can be coded directly. The simulation results show that the deterministic variables of symbolic pattern recursion graph can better describe the variation of model parameters than sort recursion graph by using Logistic map and single channel dual dynamic neuron group model. Finally, the analysis of amnesia mild cognitive impairment group and control group in actual diabetes mellitus and non-diabetes mellitus shows that the symbolic pattern recursion graph can better distinguish the deterministic variables between the two groups than the sequential recursive graph. Under the condition of cognitive state matching, it was found that diabetic patients had some relationship with cognitive function. Deterministic variables were correlated with neuropsychological tests in frontal, temporal and occipital regions.
【學(xué)位授予單位】：燕山大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2015
【分類號(hào)】：R587.2;TP391.41

【參考文獻(xiàn)】

相關(guān)期刊論文 前1條

1 謝松云;張振中;楊金孝;張坤;;腦電信號(hào)的若干處理方法研究與評(píng)價(jià)[J];計(jì)算機(jī)仿真;2007年02期

，

本文編號(hào)：1807855