Abstract: Tomodosimetry employs arrays of scintillating optical fibers to measure 2D or 3D dose distributions with high spatial resolution, making it promising for experimental validations in conformal radiotherapy. However, several aspects remain insufficiently explored, including the effects of scattered and Cherenkov radiation, the consequences of limited angular sampling, and the spatial resolution as a function of the detector design. In this work, we developed a numerical framework to evaluate the performance of a tomodosimeter under such constraints, aiming to preserve dose reconstruction quality while minimizing the number of projections. The proposed formalism incorporates both scattered and Cherenkov radiation, enabling accurate derivation of 2D dose distributions by calibrating scintillating fibers in terms of dose integral in water. Through Monte Carlo simulations of a BCF-60 fiber-based tomodosimeter, we assessed how design features, angular sampling limitations, and reconstruction methods affect the quality of reconstructed IMRT segment doses. Our results show that, even with only six projections, we achieved a 98.8% success rate of the 3%/3 mm gamma test between the reconstruction of the IMRT segment dose distribution, and the reference dose distribution. Additionally, the limiting spatial resolution of the fibers-in-contact configuration was found to be 1.53 mm, based on the 10% modulation of the oversampled modulation transfer function. These findings demonstrate that accurate and high-resolution tomodosimetric dose reconstructions are feasible with a reduced number of projections, offering a significant step forward in accelerating the clinical integration of this technology.