医学模型中的统计概念——笔记少许

supervised principal component (SPC) analysis

We propose “supervised principal component analysis (supervised
PCA)”, a generalization of PCA that is uniquely effective for regression and classification problems with high-dimensional input data. It works by estimating a sequence of principal components that have maximal dependence on the response variable. The proposed
supervised PCA is solvable in closed-form, and has a dual formulation that significantly reduces the computational complexity of problems in which the number of predictors greatly exceeds the number of observations (such as DNA microarray experiments).[1]

Cox proportional hazards model

比例风险模型——Cox回归

 Cox回归是一种允许资料有“删失（或截尾）”数据存在的，可以同时分析众多因素对生存时间影响的多变量生存分析方法。是一种半参数方法。[2]

C-index and integrated Brier scores

什么是一致性指数？

C-index，英文名全称concordance index，中文里有人翻译成一致性指数，最早是由范德堡大学（Vanderbilt University）生物统计教教授Frank E Harrell Jr 1996年提出，主要用于计算生存分析中的COX模型预测值与真实之间的区分度（discrimination），和大家熟悉的AUC其实是差不多的；在评价肿瘤患者预后模型的预测精度中用的比较多。一般评价模型的好坏主要有两个方面，一是模型的拟合优度（Goodness
of Fit),常见的评价指标主要有R方、-2logL、AIC、BIC等；另外一个是模型的预测精度，顾名思义就是模型的真实值与预测值之间差别大小，均方误差，相对误差等。在临床应用上更注重预测精度，建模的主要目的是用于预测，而C-index它就属于模型评价指标中的预测精度。

C-index的计算方法是把所研究的资料中的所有研究对象随机地两两组成对子，以生存分析为例,两个病人如果生存时间较长的一位其预测生存时间长于另一位,或预测的生存概率高的一位的生存时间长于另一位,则称之为预测结果与实际结果相符，称之为一致。[3]

integrated Brier scores -- IBS, lower scores indicating higher accuracy

log-rank test

In statistics,
the log-rank test is a hypothesis
test to compare the survival distributions
of two samples. [4]

Wilcoxon rank-sum test

In statistics,
the Mann–Whitney U test (also called the Mann–Whitney–Wilcoxon (MWW), Wilcoxon
rank-sum test, or Wilcoxon–Mann–Whitney test)
is a nonparametric test of
the null hypothesis that
it is equally likely that a randomly selected value from one sample will be less than or greater than a randomly selected value from a second sample.[5]

其它英文单词翻译及补充说明

Karnofsky performance score 卡氏评分

radiologic (Gaussian nor-malized relative cerebral blood volume and apparent diffusion co-ef cient) parameters.

参考文献

[1] Supervised principal component analysis: Visualization,
classification and regression on subspaces and submanifolds
https://www.sciencedirect.com/science/article/pii/S0031320310005819
[2] https://wenku.baidu.com/view/45349675482fb4daa58d4bc1.html

[3] https://www.shengxin.ren/article/104

[4] https://en.wikipedia.org/wiki/Log-rank_test
[5] https://en.wikipedia.org/wiki/Mann%E2%80%93Whitney_U_test