Line-scanning spectral-domain optical coherence tomography (LS-SD-OCT) is an alternative OCT configuration that enables parallel illumination and detection along the fast scanning axis, thereby eliminating the need for mechanical scanning in that direction. This feature makes LS-SD-OCT particularly suitable for high-throughput imaging applications. In such systems, the sample is illuminated by a focused line-shaped beam, which necessitates a free-space optical setup to relay the beam onto the image plane and to capture B-scans using a two-dimensional (2D) camera. However, this configuration presents inherent challenges: object points at different depths experience defocus, and scattered photons from adjacent A-scan channels can lead to crosstalk. These effects degrade resolution and reduce the clarity of both en-face and cross-sectional images. In this work, we quantitatively evaluate the imaging performance of LS-SD-OCT by constructing the point spread function (PSF) along the fast axis, taking into account both defocus and scattering in adhesive tape and onion samples. The results are systematically compared with those from free-space flying-spot SD-OCT(FS-SD-OCT) and fiber-based FS-SD-OCT systems. Our findings indicate that the image quality of LS-SD-OCT is inferior to that of FS-SD-OCT systems, offering practical insights for future enhancements of the LS-SD-OCT technique.