In this paper, we investigate the K-cell multiple input multiple output (MIMO) interference multiple-access channel (IMAC) where opportunistic user scheduling and interference alignment (IA) are combined. For traditional opportunistic IA (OIA) scheme, without consideration on enhancing the effectiveness of the desired signal, each user calculates its beamforming vector only based on the singular value decomposition (SVD) of its total leakage of interference (LIF), which leads to the performance deterioration in terms of both the achievable average sum rate and bit error rate (BER). Focusing on K-cell MIMO IMAC, we propose the signal to LIF ratio (SLIR) OIA in this paper, the beamforming vector design criterion of which takes both LIF and desired signal into consideration. Compare with the existing OIA scheme, the proposed SLIR-OIA has the same order of magnitude in computational complexity. Applying the proposed SLIR-OIA algorithm to both perfect and limited feedback scenarios, we evaluate the sum rate and BER performance through simulations, the results of which show that the proposed SLIR-OIA scheme outperforms the traditional OIA scheme significantly.