"...while elliptical galaxies are almost devoid of them." Keyword: "almost". Just for kicks and giggles I will try to search HST for any. Perhaps some of these targets are within the grasp of amateur equipment.
It took a while, but it just dawned on my that in my collection of images there is one of M31 with M32 and M110 on the same image where I combined Halpha with LRGB. So, I have imaged at least 2 elliptical galaxies in Halpha and as Wouter was already surmising, there is very little structure (i.e. none to my eye) visible in the Halpha band.
Yes, I think Wouter is correct. I've worked my way through 20 Type-E galaxies at the Hubble Legacy Archive. No H-II regions but I did find something interesting: NGC 4589 and NGC 2768 have infrared filaments in orbit around the nucleus. From what I've read, supernovae are responsible. I examined the Hubble images in various wideband filters. I plan on imaging NGC 4589 with a standard blue filter which should give good contrast. I'll let you know how it goes!