R语言，如何提取某段文字中包含某个关键字的句子？

variable 名为 ”test_result", 前三条结果如下：

[1] No recent territorial infarct, hydrocephalus, hyperdense intracranial  vessel, intracranial mass, midline shift or intracranial herniation.\r\r\n \r\r\n Small hypodensities in bilateral lentiform nuclei inferiorly are favored  to represent prominent perivascular spaces.\r\r\n Patchy low-attenuation in bilateral cerebral white matter, nonspecific but  likely represents mild chronic small vessel ischemic change. Generalized  age-related brain parenchymal volume loss

[2] Slight increase in the  ventricular volume but no transependymal hypodensity.  No new hemorrhage.   No mass effect. Midline is centered.\r\n \r\n \r\n"                                                               [3] There is increased mass effect upon the cortical  mantle but there is no midline shift.\r\n Right prefrontal subdural hematoma measures up to 3 mm in depth,  previously 4 mm in depth and has become increasingly low density.\r\n The ventricular system is unremarkable.\r\n No new hemorrhage or acute territorial ischemic insult is detected.\r\n Carotid siphon atheromatous calcification is detected.\r\n There is a left approach NG tube and paranasal sinus mucosal thickening.\r\n Unaltered calvarial fractures including bilateral petrous temporal bone  fractures resulting in fluid filling the mastoids.\r\n No abnormality of the globes/orbits is detected.\r\n \r\n \r\n"。。。。。。

（有几千个结果）                                                                     

想找出包含关键字“midline” 的句子，并产生一个新的变量（就是包含关键词的这句话），该怎么实现呢？ 感谢大神们赐教   

这样就可以了

1. new_variable <- stringr::str_subset(test_result, "midline")

复制代码

因为不是所有行都包含这个字符串，所以我这里报错了

Errors：
  replacement has 727 rows, data has 2177