Volume representation without noise and garbage

Unfortunately this solution is not for my case because my model can’t be re-segmented, it takes a long time. So I need to somehow find a way to export the clean model or to do something else. Thanks for the reply anyway.