基于静息态功能磁共振成像的戒烟结果预测研究

发布时间：2018-03-31 07:11

本文选题：静息态　切入点：功能磁共振　出处：《浙江大学》2017年硕士论文

【摘要】：目的:探讨戒烟成功者与戒烟失败者基线水平的临床资料及影像学资料间的差异,并尝试用上述差异对戒烟结果进行预测。资料与方法:78名准备戒烟者在戒烟前采集基线数据,进行T1结构像扫描以及静息态fMRI扫描。予伐尼克兰标准疗法辅助戒烟12周后,随访最终戒烟结果。提取各被试静息态fMRI数据计算ReHo、ALFF及fALFF值。随机抽取约2/3被试进行临床资料及上述影像学指标比较,提取有差异的数据纳入Logistic回归模型,对剩余被试进行戒烟结果预测,根据实际戒烟结果讨论其预测效能。结果:10名被试者因失访或扫描图像质量不合格退出研究。最终取48名被试者纳入分析,其中19人戒烟成功,29人失败。两组人群间基线临床资料及fALFF未见显著差异,ReHo及ALFF分析分别发现4个、6个有显著差异的脑区。提取相应参数分别建立Logistic回归模型,对剩余20名(其中戒烟成功者8人)被试者进行戒烟结果预测。结果显示ReHo、ALFF的预测符合率分别为70%、55%。结论:戒烟起始状态下,不同戒烟结果的吸烟者在ReHo和ALFF上存在显著差异的脑区。提取相应参数,可有效预测戒烟结果,尤其是筛选出易戒烟失败者。改变预测水平可调整预测性能,以满足不同需要。
[Abstract]:Objective: to explore the difference of clinical and imaging data between successful smokers and smokers, and try to predict the results of smoking cessation by using the above differences. Data and methods: the baseline data were collected from 78 people who were prepared to quit smoking before quitting. T1 structural images and resting fMRI scans were performed. After 12 weeks of smoking cessation, Vannickland standard therapy was used to assist in smoking cessation. The rest fMRI data were extracted to calculate the values of Rehoff and fALFF. About 2 / 3 of the subjects were randomly selected to compare the clinical data and the above imaging indexes. The different data were extracted to be included in the Logistic regression model. The results of smoking cessation were predicted for the rest of the subjects, and the predictive effectiveness was discussed according to the actual smoking cessation results. Results Ten of 10 subjects withdrew from the study because of missing interview or the quality of scanning image was not up to standard. Finally, 48 subjects were included in the analysis. Among them, 19 quit smoking successfully and 29 failed. There was no significant difference in baseline clinical data and fALFF between the two groups. Four and six brain regions with significant difference were found by ReHo and ALFF analysis. The corresponding parameters were extracted to establish Logistic regression model, respectively. The results of smoking cessation were predicted for the remaining 20 subjects (among them 8 successful smokers). The results showed that the predicted coincidence rate of ReHohal ALFF was 70% 55. Conclusion: under the initial state of smoking cessation, The brain regions of smokers with different smoking cessation results were significantly different in ReHo and ALFF. Extracting the relevant parameters could effectively predict the smoking cessation results, especially screening out the easy smokers. The predictive performance could be adjusted to meet the different needs by changing the prediction level.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R445.2;R163

【相似文献】