In theory, and I mean theory here, it is possible.
To start with, a quick review of the whole ICF concept:
The rate of fusion events is a function of the fuel's temperature and density. The former is because there is a specific energy where the reaction is maximized, and the latter because the alphas being released in those reactions provide heat to the surrounding fuel, and if that fuel is denser it doesn't have to travel as far before thermalizing.
Recall that the fuel mass is losing energy all the time through radiation, neutrons from the reactions, and other effects. So the trick to ICF is to get to a density where the alphas are heating the fuel to fusion temperatures faster than the fuel is losing energy to the environment. That is "ignition".
So what is the required density to achieve this chain reaction? About 100 times that of lead.
Now in theory, one could achieve this with any type of compression, as long as the resulting compressed mass is large enough that the alphas will stop within it, meaning that if squeezed a long cylinder into a disk, for instance, the final form has to be thick enough that an alpha travelling towards the flat sides will still stop, on average, before reaching the edge. So maybe like the aspect ratio of a quarter.
But now let us consider how we get these conditions. Lead is about 125 times the density of hydrogen gas, and we need 100 times that, so we're talking about 12,500 times compression. So if we had a cylinder of fuel 1 cm across and we wanted to compress it to 1 micron thick, it would have to start out 1.25 cm long.
Now here's where the theory gets into "nope" range. When you compress a heavier mass, like compressed hydrogen, into a lighter mass, like the remaining uncompressed fuel, bad things happen. In fact, this precise setup is the canonical example on the Wikipedia. Practically speaking, there's no way the fuel mass will ever compress down, it will fall apart at densities way lower than what you need.
So, practically, you need a sphere. A very special one, in fact, one where the fuel is deposited as a very thin layer on the inside of a denser "pusher" material so that it is collapsing into an empty center with so little material that the R-T instability is minimized. Even then, after 50 years of trying, we still can't get this to work reliably.
So when you see things like this, you can dismiss them out of hand, there is simply no way that can actually work. And more generally, non-spherical collapse is possible but will require many orders of magnitude more pusher power to make it work, to the point there is absolutely no way it will make net energy.