You will first need to get the list of series ids in the directory:
series_ids = imageReader.GetGDCMSeriesIDs(path)

Then you select the relevant series (somehow), and get the specific file names associated with it:

i=3
imageReader.GetGDCMSeriesFileNames(path, series_ids[i])

For an example with a dropbox see this Jupyter notebook.

You may be interested in the examples shown in the notebooks in general, see the main notebook repository.

Thanks a lot! It works!
BTW, here is a new question. After I read a DICOM series, I want get the meta data with it:

imageReader= sitk.ImageSeriesReader()
imageReader.LoadPrivateTagsOn()
series_ids = imageReader.GetGDCMSeriesIDs(path)
series_file_names = imageReader.GetGDCMSeriesFileNames(path, series_ids [0])
imageReader.SetFileNames(series_file_names)
image = imageReader.Execute()  # type: sitk.Image
print(image.GetMetaDataKeys())

But it returns an empty list, what should I do?
Thank you !

Three points:

1. In SimpleITK, the returned image will not have a meta data dictionary - rationale: this is most often a 3D image where each of the 2D images from which it was created had a separate meta-data dictionary. We can't know how to combine them together (there is slice specific information in each).

2. To access the meta-data keys and information on a per-slice basis you have three functions, as illustrated below:

slice_number = 0
slice_keys = imageReader.GetMetaDataKeys (slice_number) 
imageReader.HasMetaDataKey(slice_number, '0010|0010')
imageReader.GetMetaData(slice_number, '0010|0010')

3. You are missing a line before "LoadPrivateTagsOn" which tells the reader to actually load the tags:
   imageReader.MetaDataDictionaryArrayUpdateOn()

Thank you for your time. It helps a lot!
I've solved my problem.