This writeup was originally a response to a now-vanished writeup in this node. If you think this means this writeup should go away, tell me...I might agree.
I'm not a physicist, and don't even play one on TV, but this sounds a little like a linear Tokamak. Turn matter into plasma through the application of stimulus energy (RF heating, usually?) and then try to hold 'ignition,' which is the state where enough of the released fusion energy is retained in the fusion fuel mass that it becomes self-sustaining. Essentially you're trying to achieve ignition in a fuel mass with a linear escape path.
Either a leap past this or a complement to this is antimatter propulsion, in which antiparticles (or, ideally, antimolecules like antihydrogen) are reacted with 'normal' matter to produce a stream of energetic exhaust. Perhaps it might be possible to use antimatter as the stimulus for a fusion-based system; this would likely make achieving initial high energy states easier. If, however, fine control of the reaction/reacting mass is required, this may be counterproductive. Drat, I just yesterday killed the link to the antimatter interplanetary propulsion system paper...it was linked off Slashdot some months ago.
In any case, back to fusion. One big problem most Tokamaks had/have (AFAIK) is preventing the toroidal plasma produced in the reactor from expanding far enough to touch the containment wall, which instantly cools it down past fusion levels and stops the reaction. This is the same problem as trying to control the reactions mentioned earlier. Ideally, their fusing energy release wouldn't be too far above their stable-but-excited 'fuel' state, in order to ease initial ignition. If this is the case, however, it becomes easier to lose enough energy to drop below the ignition point and force a restart. If that's true, the demands on the containment/insulation/what-have-you systems become higher. If the energy release could be produced in a directed fashion, this problem might be mitigated if not avoided ("Put simply, in deference to you, Kent, it's like lasing a stick of dynamite.")
Now, there are also proposals/dreams about laser-powered spacecraft (see lightcraft, for example). The thought here is that you take a ship which has sub-critical fuel mass on board (like fusion fuel) and you use a ground-based laser to excite the fuel and either cause fusion ignition or simply cause it to turn to plasma and escape from the rear of the engine. This means you don't have to carry stuff like oxidizer, ignition systems, etc. etc. On the other hand, it means you can really only boost in one direction. These are mostly proposed for boosters designed to reach orbit. In the extreme case, you don't need fuel at all; you just use the energy from your boost laser to flash-heat atmosphere around the base of the craft, and ride the wave of expanding air upwards.
Wow, I'm waaay out off the end of the short pier of my physical science education. :-) Others, please chime in.