B-Tree index vs B-Tree as a data structure used in a storage engine?

Question

I'm reading DDIA and it references LSM Trees and B-Trees as indexes used in DBs. But it also talks about both data structures used to store data.

Are B-Trees used purely as indexes so when I query a column that I created an index on a B-Tree is used? Or are they also used to store all the data in a relational DB onto disk?

I'm a little confused as to whether these data structures have a dual purpose of being used as an index e.g. CREATE INDEX and also used as the data structures for any query I do SELECT col FROM ....