I would just get the number of decimal places and multiply it by the correct power of 10. Probably not a problem, but it would also be faster and use less memory and then flip it to a string, splitting / recombining it, and then throwing it back to double. This also works for any number of decimal places.
decimal d = 2353.61;
int count = BitConverter.GetBytes(decimal.GetBits(d)[3])[2];
d *= Math.pow(10, count);
, .