腦梗死分型與進(jìn)展性缺血性腦卒中相關(guān)性研究
發(fā)布時(shí)間：2019-06-28 19:02
【摘要】：目的探討5種腦梗死分型亞型與進(jìn)展性缺血性腦卒中的相關(guān)性；將與進(jìn)展性缺血性腦卒中相關(guān)的腦梗死分型亞型進(jìn)行聯(lián)合，構(gòu)建不同組合并相互比較，得到最準(zhǔn)確預(yù)測(cè)腦梗死進(jìn)展的最優(yōu)因素組合，為進(jìn)展性缺血性腦卒中的預(yù)防及治療提供理論依據(jù)。 方法選取經(jīng)顱腦核磁證實(shí)的急性缺血性腦卒中患者407例，，分成進(jìn)展性缺血性腦卒中組（進(jìn)展組）106例和非進(jìn)展性缺血性腦卒中組（非進(jìn)展組）301例，兩組患者均于入院時(shí)行TOAST分型、OCSP分型、CT分型、CISS分型、ASCO分型，同時(shí)收集患者的性別、年齡、吸煙史、飲酒史、高血壓病病史、糖尿病病史、冠心病病史、卒中病史等資料。對(duì)上述因素進(jìn)行單因素分析，挑選進(jìn)展性缺血性腦卒中的危險(xiǎn)因素，進(jìn)行多因素非條件Logistic回歸分析（Forward LR法）并將有統(tǒng)計(jì)學(xué)意義的因素組建不同組合。從診斷試驗(yàn)的真實(shí)性、預(yù)測(cè)效果及與原樣本吻合度3個(gè)角度對(duì)組合進(jìn)行檢驗(yàn)；最后檢驗(yàn)最優(yōu)組合的準(zhǔn)確性。 結(jié)果1經(jīng)χ2分析：高血壓病史，糖尿病病史，TOAST分型中LAA亞型、SAO亞型，OCSP分型中的TACI亞型、PACI亞型、POCI亞型、LACI亞型，CT分型中的大梗死亞型、中梗死亞型、腔隙性梗死亞型，CISS分型中的LAA亞型、PAD亞型、UE亞型，ASCO分型中的A亞型、S亞型在進(jìn)展組與非進(jìn)展組間存在統(tǒng)計(jì)學(xué)差異（P0.05）；性別、吸煙史、飲酒史、冠心病史、卒中病史及腦梗死分型中的其他亞型在進(jìn)展組與非進(jìn)展組間不存在統(tǒng)計(jì)學(xué)差異（P0.05）。經(jīng)多因素非條件Logistic回歸分析：糖尿病病史，OCSP分型中的TACI亞型、PACI亞型，CT分型中的大梗死亞型、腔隙性梗死亞型，CISS分型中UE亞型進(jìn)展組與非進(jìn)展組(組間)比較具有統(tǒng)計(jì)學(xué)意義（P0.05）。卒中進(jìn)展的風(fēng)險(xiǎn)（OR值）分別為：4.097倍、7.552倍、10.428倍、2.969倍、0.296倍、0.225倍；2對(duì)糖尿病病史，OCSP分型中的TACI亞型、PACI亞型，CT分型中的大梗死亞型、腔隙性梗死亞型，CISS分型中的UE亞型采用多因素非條件Logistic回歸分析（Forward LR法），按OR值大小依次引入，構(gòu)建不同的因素組合。通過(guò)比較組合的真實(shí)性、預(yù)測(cè)效果及與原樣本吻合度，結(jié)果顯示因素組合⑤的靈敏度(0.717)、特異度(0.841)、約登指數(shù)(0.558)、Kappa值(0.528)、陽(yáng)性預(yù)測(cè)值(0.613)、陰性預(yù)測(cè)值(0.894)較其他組合優(yōu)秀；381例患者對(duì)因素組合⑤進(jìn)行檢驗(yàn)的靈敏度=0.810，特異度=0.833，總的判對(duì)率=1－誤判率=0.827，與原樣本的吻合度Kappa=0.588。 結(jié)論1高血壓病史、糖尿病病史，TOAST分型中LAA亞型、SAO亞型，OCSP分型中的TACI亞型、PACI亞型、POCI亞型、LACI亞型，CT分型中的大梗死亞型、中梗死亞型、腔隙性梗死亞型，CISS分型中的LAA亞型、PAD亞型、UE亞型，ASCO分型中的A亞型、S亞型與進(jìn)展性缺血性腦卒中相關(guān)；2PACI亞型、TACI亞型、糖尿病病史、CISS分型中的UE亞型、CT分型中的腔隙性腦梗死亞型組合是預(yù)測(cè)進(jìn)展性缺血性腦卒中的最優(yōu)組合，能夠提高預(yù)測(cè)的準(zhǔn)確度。
[Abstract]:Objective to explore the correlation between five subtypes of cerebral infarction and progressive ischemic stroke, to combine the subtypes of cerebral infarction associated with progressive ischemic stroke, to construct different combinations and compare with each other, and to obtain the optimal factor combination to accurately predict the progress of cerebral infarction, so as to provide theoretical basis for the prevention and treatment of progressive ischemic stroke. Methods 407 patients with acute ischemic stroke confirmed by craniocerebral MRI were divided into progressive ischemic stroke group (n = 106) and non-progressive ischemic stroke group (n = 301). TOAST classification, OCSP classification, CT classification, CISS classification and ASCO classification were performed on admission. At the same time, the sex, age, smoking history, drinking history, hypertension history and diabetes history of the patients were collected. History of coronary heart disease, history of stroke, etc. The above factors were analyzed by univariate analysis, the risk factors of progressive ischemic stroke were selected, and the multivariate unconditional Logistic regression analysis (Forward LR method was used to form different combinations of statistically significant factors. The combination was tested from three angles: the authenticity of the diagnostic test, the prediction effect and the degree of coincidence with the original sample. Finally, the accuracy of the optimal combination was tested. Results 1 there were significant differences in hypertension history, diabetes history, LAA subtypes, SAO subtypes, TACI subtypes, PACI subtypes, POCI subtypes, LACI subtypes, large infarction subtypes, middle infarction subtypes, lacunar infarction subtypes, LAA subtypes, PAD subtypes, UE subtypes, A subtypes in ASCO classification between progressive group and non-progressive group (P 0.05). There was no significant difference in sex, smoking history, drinking history, coronary heart disease history, stroke history and other subtypes in cerebral infarction classification between progressive group and non-progressive group (P 0.05). Multivariate unconditional Logistic regression analysis showed that the history of diabetes mellitus, TACI subtypes in OCSP classification, PACI subtypes, large infarction subtypes in CT classification, lacunar infarction subtypes, UE subtypes in CISS classification were significantly higher than those in non-progressive groups (P 0.05). The risk of stroke progress (OR value) was 4.097 times, 7.552 times, 10.428 times, 2.969 times, 0.296 times and 0.225 times, respectively. 2 for the history of diabetes mellitus, TACI subtypes, PACI subtypes in OCSP classification, large infarction subtypes in CT classification, lacunar infarction subtypes, UE subtypes in CISS classification were introduced by multivariate unconditional Logistic regression analysis (Forward LR method), and different factor combinations were constructed according to the OR value. By comparing the authenticity, prediction effect and coincidence with the original sample, the results showed that the sensitivity (0.717), specificity (0.841), Jordan index (0.558), Kappa value (0.528), positive predictive value (0.613) and negative predictive value (0.894) of factor combination 5 were better than those of other combinations. The sensitivity of factor combination 5 was 0.810, the specificity was 0.833, the total judgment rate was 1-misjudgment rate = 0.827, and the coincidence degree Kappa=0.588. with the original sample was 0.810, 0.833, 0.827, 0.827, 0.827, 0.827, 0.827, 0.827, 0.827, 0.827, 0.827, 0.827 and 0.827, respectively. Conclusion (1) the history of hypertension, diabetes mellitus, LAA subtypes, SAO subtypes, TACI subtypes, PACI subtypes, POCI subtypes, LACI subtypes, large infarction subtypes, middle infarction subtypes, lacunar infarction subtypes, LAA subtypes, PAD subtypes, UE subtypes, A subtypes and S subtypes in CISS classification are related to progressive ischemic stroke. The combination of 2PACI subtypes, TACI subtypes, diabetes history, UE subtypes in CISS classification and lacunar cerebral infarction subtypes in CT classification is the best combination to predict progressive ischemic stroke, which can improve the accuracy of prediction.
【學(xué)位授予單位】：河北聯(lián)合大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2014
【分類(lèi)號(hào)】：R743.3
本文編號(hào)：2507532
[Abstract]:Objective to explore the correlation between five subtypes of cerebral infarction and progressive ischemic stroke, to combine the subtypes of cerebral infarction associated with progressive ischemic stroke, to construct different combinations and compare with each other, and to obtain the optimal factor combination to accurately predict the progress of cerebral infarction, so as to provide theoretical basis for the prevention and treatment of progressive ischemic stroke. Methods 407 patients with acute ischemic stroke confirmed by craniocerebral MRI were divided into progressive ischemic stroke group (n = 106) and non-progressive ischemic stroke group (n = 301). TOAST classification, OCSP classification, CT classification, CISS classification and ASCO classification were performed on admission. At the same time, the sex, age, smoking history, drinking history, hypertension history and diabetes history of the patients were collected. History of coronary heart disease, history of stroke, etc. The above factors were analyzed by univariate analysis, the risk factors of progressive ischemic stroke were selected, and the multivariate unconditional Logistic regression analysis (Forward LR method was used to form different combinations of statistically significant factors. The combination was tested from three angles: the authenticity of the diagnostic test, the prediction effect and the degree of coincidence with the original sample. Finally, the accuracy of the optimal combination was tested. Results 1 there were significant differences in hypertension history, diabetes history, LAA subtypes, SAO subtypes, TACI subtypes, PACI subtypes, POCI subtypes, LACI subtypes, large infarction subtypes, middle infarction subtypes, lacunar infarction subtypes, LAA subtypes, PAD subtypes, UE subtypes, A subtypes in ASCO classification between progressive group and non-progressive group (P 0.05). There was no significant difference in sex, smoking history, drinking history, coronary heart disease history, stroke history and other subtypes in cerebral infarction classification between progressive group and non-progressive group (P 0.05). Multivariate unconditional Logistic regression analysis showed that the history of diabetes mellitus, TACI subtypes in OCSP classification, PACI subtypes, large infarction subtypes in CT classification, lacunar infarction subtypes, UE subtypes in CISS classification were significantly higher than those in non-progressive groups (P 0.05). The risk of stroke progress (OR value) was 4.097 times, 7.552 times, 10.428 times, 2.969 times, 0.296 times and 0.225 times, respectively. 2 for the history of diabetes mellitus, TACI subtypes, PACI subtypes in OCSP classification, large infarction subtypes in CT classification, lacunar infarction subtypes, UE subtypes in CISS classification were introduced by multivariate unconditional Logistic regression analysis (Forward LR method), and different factor combinations were constructed according to the OR value. By comparing the authenticity, prediction effect and coincidence with the original sample, the results showed that the sensitivity (0.717), specificity (0.841), Jordan index (0.558), Kappa value (0.528), positive predictive value (0.613) and negative predictive value (0.894) of factor combination 5 were better than those of other combinations. The sensitivity of factor combination 5 was 0.810, the specificity was 0.833, the total judgment rate was 1-misjudgment rate = 0.827, and the coincidence degree Kappa=0.588. with the original sample was 0.810, 0.833, 0.827, 0.827, 0.827, 0.827, 0.827, 0.827, 0.827, 0.827, 0.827, 0.827 and 0.827, respectively. Conclusion (1) the history of hypertension, diabetes mellitus, LAA subtypes, SAO subtypes, TACI subtypes, PACI subtypes, POCI subtypes, LACI subtypes, large infarction subtypes, middle infarction subtypes, lacunar infarction subtypes, LAA subtypes, PAD subtypes, UE subtypes, A subtypes and S subtypes in CISS classification are related to progressive ischemic stroke. The combination of 2PACI subtypes, TACI subtypes, diabetes history, UE subtypes in CISS classification and lacunar cerebral infarction subtypes in CT classification is the best combination to predict progressive ischemic stroke, which can improve the accuracy of prediction.
【學(xué)位授予單位】：河北聯(lián)合大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2014
【分類(lèi)號(hào)】：R743.3
【參考文獻(xiàn)】
相關(guān)期刊論文 前1條
1 呂揚(yáng)勛;王圣槐;趙紅霞;;腦梗死患者血清鈣鎂含量變化及臨床意義[J];中國(guó)實(shí)用神經(jīng)疾病雜志;2006年03期
本文編號(hào)：2507532
本文鏈接：http://sikaile.net/yixuelunwen/shenjingyixue/2507532.html
最近更新
教材專(zhuān)著
