Understanding and correcting Java Application Code that is causing Stuck Threads in WebLogic

I have a production system that uses WebLogic 10.3.0.4 and this instance of WebLogic has a clustered setup (an admin server and 2 managed servers), it's running on Solaris 10.  We have an issue with the Java application (deployed as an ear) running on this instance, it gets stuck threads on the managed servers about once per week.  fyi: there is only a single application (.ear) deployed on this instance.  Our only fix at this time is to restart the managed server so that it gets back into a healthy state and removes the stuck threads.

My reason for posting today, is because I'm wanting to understand how I can better pin-point the specific cause of a stuck thread in WebLogic.  I'm aware of a few things I can do within the WebLogic Admin Console to help me get additional information on the currently stuck thread.  For instance, I know that the Admin Console allows me to help diagnose stuck threads by drilling down on the particular managed server with the stuck thread, then clicking the "thread pool" hyperlink to see data relating to the stuck thread.  I know that I can also do the "Dump Thread Stack" button to see the stack trace.  (1)