create summary results for each stage of the pipeline

[bryantChhun]

problem

currently each stage of the pipeline simply writes intermediate files with unintuitive structures. We could summarize the results in a CSV or other standard text-format.

For example:

run_segmentation.py --method instance_segmentation

could return a metadata txt file that describes the number of cell positions per timepoint.

Another example:

run_patch.py --method build_trajectories

could return the number of trajectories, positions, or other statistics, in a text file.