Well a transdermal patch would be some type of gel (like fentanyl patches) but I honestly hadn't even thought of that.
One key aspect of transdermal success would be applying the DMSO/DMT(acid salt) solution to an area that is high in blood flow, such as the radial artery. Because of the thin skin and the rushing river of blood, I don't see why it wouldn't work.
After efforts had been made to establish the viability of transfer, you could narrow the range of effective dose, starting like Strassman and working up to 0.4mg/kg.
One reason this method failed in the past is because it was attempted by smearing the solution on the chest. Though the solution will go into many blood vessels here, it lacks a major arterial pathway that is not protected by, say the ribcage (a la aorta).
@Guyomech: I believe this ROA would be similar to IV injection, with a much softer, more subtle and elaborate come up but similar duration. Think about the DMT entering the bloodstream, it's coming through the skin (indirectly) instead of being forced into the vein.
I have an interest in exploring an extended intravenous injection that continuously drips DMT into a subject at the metabolic rate of consumption, thus keeping the DMT in their blood at a certain level. The idea for this would be to extend the "hyperspace" time, or the time the subject's ego was completely dissolved to better understand the effects. However, I don't know if your body could maintain that state without failing.
"Thoughts come from somewhere. That's where I want to be." -Apis mellifera