Here's a general procedure you can use which assumes that you original
regions had matching boundaaries.
1) Convert to Polylines as you did.
2) Combine all these polylines into one big polyline. This process
will remove the duplicate boundary sections. (Objects Combine)
3) Disaggregate the big polyline object to individual sectional
polylines (Objects Disaggregate).
You should now have a set of polylines where each polyline represents
an original boundary section from the original data set.
Eric Blasenheim
Pitney Bowes Software.