As the high-frequency analog to field-scale earthquakes, acoustic emissions (AE) provide valuable complement to study rock deformation mechanisms. During our latest load-stepping creep experiments with CO2 and water injection into a basaltic sample from Carbfix site in Iceland, ~8,800 AE events are detected by at least one of the seven Piezoelectric sensors. Here, we apply a cross-correlation-based source imaging method, called geometric-mean reverse-time migration (GmRTM) to locate those AE events. Besides the attractive picking-free feature shared with other waveform-based methods, GmRTM is advantageous in generating higher-resolution source images with fewer imaging artifacts, compared to the conventional time-reversal imaging (TRI) or back-projection method, especially with sparse receivers as in most AE monitoring settings. In addition, GmRTM circumvents the non-trivial onset time estimation of TRI, which often adds on the location uncertainty. The imaged AE events are generally scattered across the sample, indicating a complicated fracture network rather than a well-defined major shear fracture plane. In addition to the benchmark test with our X-ray computed tomography results, the location accuracy is also confirmed by the overall coherence of waveforms of selected events, which are shifted according to the calculated traveltimes. We additionally conduct a jackknife-style validation by checking the consistency of the imaging results when rejecting data from the noisiest channel. Next, we select ~1500 AE events with high SNR and conduct moment tensor estimation using back-propagated stress tensor at the tempo-spatial locations. Approximating the inverse operation with its adjoint counterpart (i.e., back-propagation) can generate more stable results with limited receiver numbers. The estimated AE locations and the corresponding focal mechanisms supports our previously inferred micro-fracking mechanism during transient creep deformation.