Using event biase techniques to increase the counts of detector

Dear Geant4 experts:
I want to simulate a X-ray CT system and record the photon energy spectrum in the detector.
The primary source is a cone beam with determined energy distribution. The detector is a matrix (512 x 2048), each cell has a demension of 0.1 x 0.1 x 1.0 mm3 (facing source is 0.1 x 0.1 mm2). A phantom located between the source and detector. The energy of photon reachs the detector is recorded. The geometry is shown in the attach file.
I simulated 10^9 events in a run. However, I found the photons reache the detector cell is very low (about 50 - 200 counts with phantom, and 300 counts without). To obtain better statistical results, it is desirable to simulate a larger number of particles.
My question is that could some event biase techniques can be used in my simulation? (My concern is the number and energy of particles reaching the detector´╝î not the dose deposition)