Oblique lateral interbody fusion (OLIF) surgery is a minimally invasive spinal surgery technique that has become increasingly popular in recent years. The primary objective of the current study was to design a minimally invasive expandable fusion device that can reduce iatrogenic nerve damage and minimize endplate damage during OLIF surgery, while restoring intervertebral height and alignment. The second objective was to use finite element analysis to evaluate the biomechanical stability of the newly designed expandable fusion device after implantation into the intervertebral space. A new bidirectional expandable cage was designed in this study. A finite element model (FEM) of L3-L5 lumbar segment was modified to simulate decompression and fusion. The modified FEMs were constructed in the following cases: intact model, bidirectional expandable cage (alone, with unilateral pedicle screws [UPSs], and with bilateral pedicle screws [BPSs]) model, conventional OLIF cage (alone, with UPSs, and with BPSs) model. To simulate physiological loadings, the models were subjected to a follower compressive pre-load of 400 N, in addition to 8.0 Nm of flexion, extension, lateral bending, and axial rotation moments. All modified FEMs exhibited a significant reduction in motion at L3-L5 compared to the intact model. Among the fusion models, the bidirectional expandable cage (BEC) with BPS model displayed the highest stiffness and demonstrated a reduced range of motion (48.5%-75.7%). Additionally, the peak stress on the endplate in the conventional OLIF cage (Conv-OLIF) model was generally lower than that in the BEC models. The cage in the BEC ALONE model exhibited the highest stress (93.87-176.3 MPa) on the endplate in most motion modes, while the cage in the Conv-OLIF+BPS model had the lowest stress (16.67-30.58 MPa) on the endplate in most motion modes. The maximum stress on the fixation in the BEC fusion models was generally lower than that in the Conv-OLIF fusion group under the same loading conditions. The OLIF ALONE model had the lowest stress on the adjacent disc, while the stress level in the BEC ALONE model was very close to it. The BEC implanted models had higher stiffness, and more proper stress distribution on the posterior fixation was comparable to that of the Conv-OLIF models. However, the endplate stress peaks and cage stress peaks of the BEC models were slightly higher than those of the Conv-OLIF models, though still within a clinically acceptable range. Taking into account both biomechanical and clinical perspectives, BEC-assisted unilateral pedicle screw fixation meet clinical demand and may serve as a viable alternative to Conv-OLIF fusion.